02_notebook_executed.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6a8d0182",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Feature engineering on extracted Pull Requests data\n",
    "\n",
    "\n",
    "## What we did previously\n",
    "\n",
    "In the previous step, we collected data from a Github repository of choice and store them on S3 storage\n",
    "\n",
    "## In this step\n",
    "\n",
    "In this notebook we will walk through how to start analyzing the PR data collected from a Github repository.\n",
    "\n",
    "We explore the various fields in the PR dataset and the features that we can derive from them in order to frame an a Machine Learning where we predict the time taken to merge a PR."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "dfe11a56",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:37.636067Z",
     "iopub.status.busy": "2022-12-07T15:27:37.633347Z",
     "iopub.status.idle": "2022-12-07T15:27:39.674758Z",
     "shell.execute_reply": "2022-12-07T15:27:39.673976Z"
    }
   },
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "import gzip\n",
    "import json\n",
    "import boto3\n",
    "import pathlib\n",
    "import datetime\n",
    "from tqdm import tqdm\n",
    "from io import BytesIO\n",
    "from pathlib import Path\n",
    "from functools import partial\n",
    "from collections import Counter\n",
    "from github import Github\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from wordcloud import STOPWORDS\n",
    "\n",
    "from sklearn.feature_selection import SelectKBest\n",
    "from sklearn.feature_selection import chi2\n",
    "from sklearn.feature_selection import mutual_info_classif, f_classif\n",
    "\n",
    "from dotenv import load_dotenv, find_dotenv\n",
    "import ceph_comm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9476e951",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:39.680967Z",
     "iopub.status.busy": "2022-12-07T15:27:39.679463Z",
     "iopub.status.idle": "2022-12-07T15:27:39.696161Z",
     "shell.execute_reply": "2022-12-07T15:27:39.695466Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sns.set(rc={\"figure.figsize\": (20, 10)})\n",
    "load_dotenv(find_dotenv(), override=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9bc3a2c6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:39.704785Z",
     "iopub.status.busy": "2022-12-07T15:27:39.703421Z",
     "iopub.status.idle": "2022-12-07T15:27:39.714757Z",
     "shell.execute_reply": "2022-12-07T15:27:39.714128Z"
    }
   },
   "outputs": [],
   "source": [
    "## CEPH Bucket variables\n",
    "## Create a .env file on your local with the correct configs,\n",
    "\n",
    "ORG = os.getenv(\"GITHUB_ORG\")\n",
    "REPO = os.getenv(\"GITHUB_REPO\")\n",
    "\n",
    "## S3 bucket credentials\n",
    "s3_endpoint_url = os.getenv(\"S3_ENDPOINT_URL\")\n",
    "s3_access_key = os.getenv(\"S3_ACCESS_KEY\")\n",
    "s3_secret_key = os.getenv(\"S3_SECRET_KEY\")\n",
    "s3_bucket = os.getenv(\"S3_BUCKET\")\n",
    "\n",
    "REMOTE = os.getenv(\"REMOTE\")\n",
    "\n",
    "s3_input_data_path = os.getenv(\"CEPH_BUCKET_PREFIX\")\n",
    "\n",
    "RAW_DATA_PATH = os.path.join(s3_input_data_path, ORG, REPO)\n",
    "TOKEN = os.getenv(\"GITHUB_ACCESS_TOKEN\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "38f32d2b-22dc-419b-b76a-ce41e35ef58e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:39.720278Z",
     "iopub.status.busy": "2022-12-07T15:27:39.718601Z",
     "iopub.status.idle": "2022-12-07T15:27:45.508702Z",
     "shell.execute_reply": "2022-12-07T15:27:45.507811Z"
    }
   },
   "outputs": [],
   "source": [
    "# Collect PRs and combine them\n",
    "cc = ceph_comm.CephCommunication(s3_endpoint_url, s3_access_key, s3_secret_key, s3_bucket)\n",
    "\n",
    "# how many files of the given extension were changed in PR\n",
    "CLOSED_PR_IDS_FILENAME = os.path.join(ORG + REPO + \"CLOSED_PR_IDS.parquet\")\n",
    "CLOSED_PRS_KEY = os.path.join(s3_input_data_path, ORG, REPO, \"closed_prs\")\n",
    "PR_ID_DF = cc.read_from_ceph(CLOSED_PRS_KEY, CLOSED_PR_IDS_FILENAME)\n",
    "pr_ids = list(PR_ID_DF['closed_pr_ids'])\n",
    "\n",
    "REPO_PATH = os.path.join(s3_input_data_path, ORG, REPO)\n",
    "\n",
    "appended_prs = []\n",
    "for pr_id in pr_ids:\n",
    "    PR_PATH = os.path.join(REPO_PATH, \"PRs/\"+ str(pr_id) + \".json\")\n",
    "    pr = cc.get_json_from_ceph(PR_PATH)\n",
    "    appended_prs.append(pr)\n",
    "\n",
    "appended_prs = pd.concat(appended_prs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "115b9040-4b29-4445-b34a-3c1183707464",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:45.513015Z",
     "iopub.status.busy": "2022-12-07T15:27:45.512445Z",
     "iopub.status.idle": "2022-12-07T15:27:45.540652Z",
     "shell.execute_reply": "2022-12-07T15:27:45.539738Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>body</th>\n",
       "      <th>size</th>\n",
       "      <th>created_by</th>\n",
       "      <th>created_at</th>\n",
       "      <th>closed_at</th>\n",
       "      <th>closed_by</th>\n",
       "      <th>merged_at</th>\n",
       "      <th>merged_by</th>\n",
       "      <th>commits_number</th>\n",
       "      <th>changed_files_number</th>\n",
       "      <th>interactions</th>\n",
       "      <th>reviews</th>\n",
       "      <th>labels</th>\n",
       "      <th>commits</th>\n",
       "      <th>changed_files</th>\n",
       "      <th>first_review_at</th>\n",
       "      <th>first_approve_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>update inference output text</td>\n",
       "      <td>Updated `03_model_training.ipynb` to create a ...</td>\n",
       "      <td>XXL</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1670362761</td>\n",
       "      <td>1670426174</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1670426174</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>{'github-actions[bot]': 20, 'MichaelClifford':...</td>\n",
       "      <td>{'1207369509': {'author': 'oindrillac', 'words...</td>\n",
       "      <td>[]</td>\n",
       "      <td>[e2bfe03be2fae4095ec22c604c876ae53c4f4c4a]</td>\n",
       "      <td>[03_model_training.ipynb, 04_model_inference.i...</td>\n",
       "      <td>1670362850</td>\n",
       "      <td>1670367056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Update inference.yaml.</td>\n",
       "      <td>test</td>\n",
       "      <td>S</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1670273067</td>\n",
       "      <td>1670276461</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1670276461</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>{'github-actions[bot]': 40}</td>\n",
       "      <td>{'1205314017': {'author': 'MichaelClifford', '...</td>\n",
       "      <td>[]</td>\n",
       "      <td>[1a92e5f7feba314360bffc318d88d89e6decb4b0, 57a...</td>\n",
       "      <td>[.github/workflows/inference.yaml, action.yml]</td>\n",
       "      <td>1670275829</td>\n",
       "      <td>1670275829</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>testing 'on PR' functionality. Do Not Merge</td>\n",
       "      <td>:wave:</td>\n",
       "      <td>XS</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1670272320</td>\n",
       "      <td>1670276415</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>{'github-actions[bot]': 10}</td>\n",
       "      <td>{}</td>\n",
       "      <td>[]</td>\n",
       "      <td>[24c6e24399b482bf9f4e7cd5bd94ed418ca919e3]</td>\n",
       "      <td>[test_file.txt]</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Add `with: MODE` functionality</td>\n",
       "      <td>Addresses #32</td>\n",
       "      <td>S</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1670259791</td>\n",
       "      <td>1670273808</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>{'github-actions[bot]': 10}</td>\n",
       "      <td>{}</td>\n",
       "      <td>[]</td>\n",
       "      <td>[b00830414636895c2790219c0a4b8b6fb812b44f, 705...</td>\n",
       "      <td>[.github/workflows/inference.yaml, action.yml]</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>fix how inference was callingPRs</td>\n",
       "      <td>There was a bug in how the inference notebook ...</td>\n",
       "      <td>L</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1669906871</td>\n",
       "      <td>1669907261</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1669907261</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>{}</td>\n",
       "      <td>{'1201160046': {'author': 'oindrillac', 'words...</td>\n",
       "      <td>[]</td>\n",
       "      <td>[f8d860b1f6f4cf770293e5d377bb2bb37b996102]</td>\n",
       "      <td>[04_model_inference.ipynb]</td>\n",
       "      <td>1669907047</td>\n",
       "      <td>1669907047</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         title  \\\n",
       "0                 update inference output text   \n",
       "0                      Update inference.yaml.    \n",
       "0  testing 'on PR' functionality. Do Not Merge   \n",
       "0              Add `with: MODE` functionality    \n",
       "0             fix how inference was callingPRs   \n",
       "\n",
       "                                                body size       created_by  \\\n",
       "0  Updated `03_model_training.ipynb` to create a ...  XXL  MichaelClifford   \n",
       "0                                              test     S  aakankshaduggal   \n",
       "0                                            :wave:    XS  MichaelClifford   \n",
       "0                                     Addresses #32     S  aakankshaduggal   \n",
       "0  There was a bug in how the inference notebook ...    L  MichaelClifford   \n",
       "\n",
       "   created_at   closed_at        closed_by   merged_at        merged_by  \\\n",
       "0  1670362761  1670426174  MichaelClifford  1670426174  MichaelClifford   \n",
       "0  1670273067  1670276461  MichaelClifford  1670276461  MichaelClifford   \n",
       "0  1670272320  1670276415  MichaelClifford        None             None   \n",
       "0  1670259791  1670273808  aakankshaduggal        None             None   \n",
       "0  1669906871  1669907261  aakankshaduggal  1669907261  aakankshaduggal   \n",
       "\n",
       "   commits_number  changed_files_number  \\\n",
       "0               1                     2   \n",
       "0               4                     2   \n",
       "0               1                     1   \n",
       "0               3                     2   \n",
       "0               1                     1   \n",
       "\n",
       "                                        interactions  \\\n",
       "0  {'github-actions[bot]': 20, 'MichaelClifford':...   \n",
       "0                        {'github-actions[bot]': 40}   \n",
       "0                        {'github-actions[bot]': 10}   \n",
       "0                        {'github-actions[bot]': 10}   \n",
       "0                                                 {}   \n",
       "\n",
       "                                             reviews labels  \\\n",
       "0  {'1207369509': {'author': 'oindrillac', 'words...     []   \n",
       "0  {'1205314017': {'author': 'MichaelClifford', '...     []   \n",
       "0                                                 {}     []   \n",
       "0                                                 {}     []   \n",
       "0  {'1201160046': {'author': 'oindrillac', 'words...     []   \n",
       "\n",
       "                                             commits  \\\n",
       "0         [e2bfe03be2fae4095ec22c604c876ae53c4f4c4a]   \n",
       "0  [1a92e5f7feba314360bffc318d88d89e6decb4b0, 57a...   \n",
       "0         [24c6e24399b482bf9f4e7cd5bd94ed418ca919e3]   \n",
       "0  [b00830414636895c2790219c0a4b8b6fb812b44f, 705...   \n",
       "0         [f8d860b1f6f4cf770293e5d377bb2bb37b996102]   \n",
       "\n",
       "                                       changed_files first_review_at  \\\n",
       "0  [03_model_training.ipynb, 04_model_inference.i...      1670362850   \n",
       "0     [.github/workflows/inference.yaml, action.yml]      1670275829   \n",
       "0                                    [test_file.txt]            None   \n",
       "0     [.github/workflows/inference.yaml, action.yml]            None   \n",
       "0                         [04_model_inference.ipynb]      1669907047   \n",
       "\n",
       "  first_approve_at  \n",
       "0       1670367056  \n",
       "0       1670275829  \n",
       "0             None  \n",
       "0             None  \n",
       "0       1669907047  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "appended_prs.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "d02d7136-2bb4-4ec8-903c-315afb22c524",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:45.544647Z",
     "iopub.status.busy": "2022-12-07T15:27:45.543991Z",
     "iopub.status.idle": "2022-12-07T15:27:45.928742Z",
     "shell.execute_reply": "2022-12-07T15:27:45.927644Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ResponseMetadata': {'RequestId': 'lbdt0yim-auwjbg-off',\n",
       "  'HostId': 'lbdt0yim-auwjbg-off',\n",
       "  'HTTPStatusCode': 200,\n",
       "  'HTTPHeaders': {'x-amz-request-id': 'lbdt0yim-auwjbg-off',\n",
       "   'x-amz-id-2': 'lbdt0yim-auwjbg-off',\n",
       "   'access-control-allow-origin': '*',\n",
       "   'access-control-allow-credentials': 'true',\n",
       "   'access-control-allow-methods': 'GET,POST,PUT,DELETE,OPTIONS',\n",
       "   'access-control-allow-headers': 'Content-Type,Content-MD5,Authorization,X-Amz-User-Agent,X-Amz-Date,ETag,X-Amz-Content-Sha256',\n",
       "   'access-control-expose-headers': 'ETag,X-Amz-Version-Id',\n",
       "   'etag': '\"f9e31c14d53d2aca9fcb24c1b7ce5f11\"',\n",
       "   'date': 'Wed, 07 Dec 2022 15:27:45 GMT',\n",
       "   'keep-alive': 'timeout=5',\n",
       "   'content-length': '0',\n",
       "   'set-cookie': '1a4aa612fe797ac8466d7ee00e5520d5=b9a4e3efe932781180a9a8945a639f83; path=/; HttpOnly; Secure; SameSite=None'},\n",
       "  'RetryAttempts': 0},\n",
       " 'ETag': '\"f9e31c14d53d2aca9fcb24c1b7ce5f11\"'}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cc.upload_to_ceph_as_json(appended_prs, RAW_DATA_PATH, \"PullRequest.json\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "4d1b0d9f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:45.933378Z",
     "iopub.status.busy": "2022-12-07T15:27:45.932603Z",
     "iopub.status.idle": "2022-12-07T15:27:46.161028Z",
     "shell.execute_reply": "2022-12-07T15:27:46.160055Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "getting dataset from ceph\n"
     ]
    }
   ],
   "source": [
    "print(\"getting dataset from ceph\")\n",
    "pr_df = cc.get_json_from_ceph(os.path.join(RAW_DATA_PATH, \"PullRequest.json\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "70c24185-4fbf-4287-98fd-c8d721ef71fa",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.165146Z",
     "iopub.status.busy": "2022-12-07T15:27:46.164653Z",
     "iopub.status.idle": "2022-12-07T15:27:46.193997Z",
     "shell.execute_reply": "2022-12-07T15:27:46.193104Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>body</th>\n",
       "      <th>size</th>\n",
       "      <th>created_by</th>\n",
       "      <th>created_at</th>\n",
       "      <th>closed_at</th>\n",
       "      <th>closed_by</th>\n",
       "      <th>merged_at</th>\n",
       "      <th>merged_by</th>\n",
       "      <th>commits_number</th>\n",
       "      <th>changed_files_number</th>\n",
       "      <th>interactions</th>\n",
       "      <th>reviews</th>\n",
       "      <th>labels</th>\n",
       "      <th>commits</th>\n",
       "      <th>changed_files</th>\n",
       "      <th>first_review_at</th>\n",
       "      <th>first_approve_at</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>update inference output text</td>\n",
       "      <td>Updated `03_model_training.ipynb` to create a ...</td>\n",
       "      <td>XXL</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1670362761</td>\n",
       "      <td>1670426174</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1.670426e+09</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>{'github-actions[bot]': 20, 'MichaelClifford':...</td>\n",
       "      <td>{'1207369509': {'author': 'oindrillac', 'words...</td>\n",
       "      <td>[]</td>\n",
       "      <td>[e2bfe03be2fae4095ec22c604c876ae53c4f4c4a]</td>\n",
       "      <td>[03_model_training.ipynb, 04_model_inference.i...</td>\n",
       "      <td>1.670363e+09</td>\n",
       "      <td>1.670367e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Update inference.yaml.</td>\n",
       "      <td>test</td>\n",
       "      <td>S</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1670273067</td>\n",
       "      <td>1670276461</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1.670276e+09</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>4</td>\n",
       "      <td>2</td>\n",
       "      <td>{'github-actions[bot]': 40}</td>\n",
       "      <td>{'1205314017': {'author': 'MichaelClifford', '...</td>\n",
       "      <td>[]</td>\n",
       "      <td>[1a92e5f7feba314360bffc318d88d89e6decb4b0, 57a...</td>\n",
       "      <td>[.github/workflows/inference.yaml, action.yml]</td>\n",
       "      <td>1.670276e+09</td>\n",
       "      <td>1.670276e+09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>testing 'on PR' functionality. Do Not Merge</td>\n",
       "      <td>:wave:</td>\n",
       "      <td>XS</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1670272320</td>\n",
       "      <td>1670276415</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>{'github-actions[bot]': 10}</td>\n",
       "      <td>{}</td>\n",
       "      <td>[]</td>\n",
       "      <td>[24c6e24399b482bf9f4e7cd5bd94ed418ca919e3]</td>\n",
       "      <td>[test_file.txt]</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Add `with: MODE` functionality</td>\n",
       "      <td>Addresses #32</td>\n",
       "      <td>S</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1670259791</td>\n",
       "      <td>1670273808</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>NaN</td>\n",
       "      <td>None</td>\n",
       "      <td>3</td>\n",
       "      <td>2</td>\n",
       "      <td>{'github-actions[bot]': 10}</td>\n",
       "      <td>{}</td>\n",
       "      <td>[]</td>\n",
       "      <td>[b00830414636895c2790219c0a4b8b6fb812b44f, 705...</td>\n",
       "      <td>[.github/workflows/inference.yaml, action.yml]</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>fix how inference was callingPRs</td>\n",
       "      <td>There was a bug in how the inference notebook ...</td>\n",
       "      <td>L</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1669906871</td>\n",
       "      <td>1669907261</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1.669907e+09</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>{}</td>\n",
       "      <td>{'1201160046': {'author': 'oindrillac', 'words...</td>\n",
       "      <td>[]</td>\n",
       "      <td>[f8d860b1f6f4cf770293e5d377bb2bb37b996102]</td>\n",
       "      <td>[04_model_inference.ipynb]</td>\n",
       "      <td>1.669907e+09</td>\n",
       "      <td>1.669907e+09</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         title  \\\n",
       "0                 update inference output text   \n",
       "1                      Update inference.yaml.    \n",
       "2  testing 'on PR' functionality. Do Not Merge   \n",
       "3              Add `with: MODE` functionality    \n",
       "4             fix how inference was callingPRs   \n",
       "\n",
       "                                                body size       created_by  \\\n",
       "0  Updated `03_model_training.ipynb` to create a ...  XXL  MichaelClifford   \n",
       "1                                              test     S  aakankshaduggal   \n",
       "2                                            :wave:    XS  MichaelClifford   \n",
       "3                                     Addresses #32     S  aakankshaduggal   \n",
       "4  There was a bug in how the inference notebook ...    L  MichaelClifford   \n",
       "\n",
       "   created_at   closed_at        closed_by     merged_at        merged_by  \\\n",
       "0  1670362761  1670426174  MichaelClifford  1.670426e+09  MichaelClifford   \n",
       "1  1670273067  1670276461  MichaelClifford  1.670276e+09  MichaelClifford   \n",
       "2  1670272320  1670276415  MichaelClifford           NaN             None   \n",
       "3  1670259791  1670273808  aakankshaduggal           NaN             None   \n",
       "4  1669906871  1669907261  aakankshaduggal  1.669907e+09  aakankshaduggal   \n",
       "\n",
       "   commits_number  changed_files_number  \\\n",
       "0               1                     2   \n",
       "1               4                     2   \n",
       "2               1                     1   \n",
       "3               3                     2   \n",
       "4               1                     1   \n",
       "\n",
       "                                        interactions  \\\n",
       "0  {'github-actions[bot]': 20, 'MichaelClifford':...   \n",
       "1                        {'github-actions[bot]': 40}   \n",
       "2                        {'github-actions[bot]': 10}   \n",
       "3                        {'github-actions[bot]': 10}   \n",
       "4                                                 {}   \n",
       "\n",
       "                                             reviews labels  \\\n",
       "0  {'1207369509': {'author': 'oindrillac', 'words...     []   \n",
       "1  {'1205314017': {'author': 'MichaelClifford', '...     []   \n",
       "2                                                 {}     []   \n",
       "3                                                 {}     []   \n",
       "4  {'1201160046': {'author': 'oindrillac', 'words...     []   \n",
       "\n",
       "                                             commits  \\\n",
       "0         [e2bfe03be2fae4095ec22c604c876ae53c4f4c4a]   \n",
       "1  [1a92e5f7feba314360bffc318d88d89e6decb4b0, 57a...   \n",
       "2         [24c6e24399b482bf9f4e7cd5bd94ed418ca919e3]   \n",
       "3  [b00830414636895c2790219c0a4b8b6fb812b44f, 705...   \n",
       "4         [f8d860b1f6f4cf770293e5d377bb2bb37b996102]   \n",
       "\n",
       "                                       changed_files  first_review_at  \\\n",
       "0  [03_model_training.ipynb, 04_model_inference.i...     1.670363e+09   \n",
       "1     [.github/workflows/inference.yaml, action.yml]     1.670276e+09   \n",
       "2                                    [test_file.txt]              NaN   \n",
       "3     [.github/workflows/inference.yaml, action.yml]              NaN   \n",
       "4                         [04_model_inference.ipynb]     1.669907e+09   \n",
       "\n",
       "   first_approve_at  \n",
       "0      1.670367e+09  \n",
       "1      1.670276e+09  \n",
       "2               NaN  \n",
       "3               NaN  \n",
       "4      1.669907e+09  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pr_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "13dc895b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.198018Z",
     "iopub.status.busy": "2022-12-07T15:27:46.197372Z",
     "iopub.status.idle": "2022-12-07T15:27:46.202466Z",
     "shell.execute_reply": "2022-12-07T15:27:46.201698Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(27, 18)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pr_df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c472ffbd",
   "metadata": {},
   "source": [
    "## Feature engineering\n",
    "\n",
    "In the next section we will do some basic feature engineering and correlation analyis on our dataset to see if anything interesting jumps out at us. "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4ed68d9a",
   "metadata": {},
   "source": [
    "**Time to Merge**\n",
    "\n",
    "We will derive a new feature called `time_to_merge` which will be the difference between the `created_at` and `merged_at` fields. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "d67035e7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.207249Z",
     "iopub.status.busy": "2022-12-07T15:27:46.206508Z",
     "iopub.status.idle": "2022-12-07T15:27:46.212376Z",
     "shell.execute_reply": "2022-12-07T15:27:46.211594Z"
    }
   },
   "outputs": [],
   "source": [
    "# derive time_to_merege\n",
    "pr_df[\"time_to_merge\"] = pr_df[\"merged_at\"] - pr_df[\"created_at\"]\n",
    "pr_df[\"time_to_merge\"] = pr_df[\"time_to_merge\"].astype(\"float\")\n",
    "# NOTE: we cant fill NA with 0 because that would imply the PR got merged instantaneously"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "fddf9750-49f5-4997-8280-e94e1e778ad9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.215894Z",
     "iopub.status.busy": "2022-12-07T15:27:46.215259Z",
     "iopub.status.idle": "2022-12-07T15:27:46.222151Z",
     "shell.execute_reply": "2022-12-07T15:27:46.221389Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      63413.0\n",
       "1       3394.0\n",
       "2          NaN\n",
       "3          NaN\n",
       "4        390.0\n",
       "5         93.0\n",
       "6       1355.0\n",
       "7          NaN\n",
       "8        229.0\n",
       "9        309.0\n",
       "10      5860.0\n",
       "11     24070.0\n",
       "12    240653.0\n",
       "13      5822.0\n",
       "14         NaN\n",
       "15      9962.0\n",
       "16    324073.0\n",
       "17      2077.0\n",
       "18      7678.0\n",
       "19      3819.0\n",
       "20         NaN\n",
       "21         NaN\n",
       "22         NaN\n",
       "23    101246.0\n",
       "24     65665.0\n",
       "25      2713.0\n",
       "26    937820.0\n",
       "Name: time_to_merge, dtype: float64"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pr_df[\"time_to_merge\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "939d196e",
   "metadata": {},
   "source": [
    "Let's look at the distribution of the time to merge column. This can help us determine what the setup for ML problems such as [#236](https://github.com/aicoe-aiops/ocp-ci-analysis/issues/236) could be like. That is, should the time to merge problem be framed as a regression or classification problem.\n",
    "\n",
    "First, lets try to determine what the granularity of the output values should be, in case of a regression setup. That is, should the model predict time to merge in number of seconds, number of hours, number of days, or some other chunk of time."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "b8cc628b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.226742Z",
     "iopub.status.busy": "2022-12-07T15:27:46.226104Z",
     "iopub.status.idle": "2022-12-07T15:27:46.234489Z",
     "shell.execute_reply": "2022-12-07T15:27:46.233748Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0     0.80\n",
       "3.0     0.05\n",
       "4.0     0.05\n",
       "2.0     0.05\n",
       "11.0    0.05\n",
       "Name: time_to_merge, dtype: float64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# what does the distribution of values look like if we choose days as the granularity\n",
    "ttm_days = np.ceil(pr_df.time_to_merge / (60 * 60 * 24))\n",
    "\n",
    "# value counts - to how many PRs have value 1, 2, 3, etc. for days to merge\n",
    "vc = ttm_days.value_counts()\n",
    "\n",
    "# normalized value counts (i.e. percent of PRs instead of raw number)\n",
    "nvc = vc / vc.sum()\n",
    "nvc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "192aeb22",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.238959Z",
     "iopub.status.busy": "2022-12-07T15:27:46.238296Z",
     "iopub.status.idle": "2022-12-07T15:27:46.247029Z",
     "shell.execute_reply": "2022-12-07T15:27:46.246280Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0     0.65\n",
       "6.0     0.05\n",
       "3.0     0.05\n",
       "23.0    0.05\n",
       "31.0    0.05\n",
       "10.0    0.05\n",
       "7.0     0.05\n",
       "87.0    0.05\n",
       "Name: time_to_merge, dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# what if we chose some intermediate value e.g. 3 hours as the granularity?\n",
    "ttm_3hours = np.ceil(pr_df.time_to_merge / (60 * 60 * 3))\n",
    "\n",
    "# value counts - to how many PRs have value 1, 2, 3, etc. for num. 3-hours to merge\n",
    "vc = ttm_3hours.value_counts()\n",
    "\n",
    "# normalized value counts (i.e. percent of PRs instead of raw number)\n",
    "nvc = vc / vc.sum()\n",
    "nvc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "44e09f11",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.251713Z",
     "iopub.status.busy": "2022-12-07T15:27:46.250959Z",
     "iopub.status.idle": "2022-12-07T15:27:46.259585Z",
     "shell.execute_reply": "2022-12-07T15:27:46.258811Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0      0.40\n",
       "2.0      0.15\n",
       "3.0      0.10\n",
       "18.0     0.05\n",
       "7.0      0.05\n",
       "67.0     0.05\n",
       "91.0     0.05\n",
       "29.0     0.05\n",
       "19.0     0.05\n",
       "261.0    0.05\n",
       "Name: time_to_merge, dtype: float64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# what if we chose hours as the granularity?\n",
    "ttm_hours = np.ceil(pr_df.time_to_merge / (60 * 60))\n",
    "\n",
    "# value counts - to how many PRs have value 1, 2, 3, etc. for hours to merge\n",
    "vc = ttm_hours.value_counts()\n",
    "\n",
    "# normalized value counts (i.e. percent of PRs instead of raw number)\n",
    "nvc = vc / vc.sum()\n",
    "nvc"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fb8e316b",
   "metadata": {},
   "source": [
    "Now, lets try to determine, if we were to set this up as a classification problem, what the output classes should be. In the following cell, we'll try to split the time-to-merge values into 10 equally populated buckets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "0ca35bc5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.263572Z",
     "iopub.status.busy": "2022-12-07T15:27:46.262922Z",
     "iopub.status.idle": "2022-12-07T15:27:46.271409Z",
     "shell.execute_reply": "2022-12-07T15:27:46.270631Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0     0.025833\n",
       "0.1     0.083611\n",
       "0.2     0.322778\n",
       "0.3     0.700611\n",
       "0.4     1.013611\n",
       "0.5     1.622500\n",
       "0.6     2.386556\n",
       "0.7     9.964694\n",
       "0.8    20.217000\n",
       "0.9    69.165278\n",
       "Name: time_to_merge, dtype: float64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# lets look at what the 10th, 20th, 30th etc percentiles are, to see what the bin boundaries (in terms of hours)\n",
    "# would be if we were to split the data in 10 equally populated bins\n",
    "n_buckets = 10\n",
    "\n",
    "quantiles = pr_df.time_to_merge.quantile(q=np.arange(0, 1 + 1e-100, 1 / n_buckets))\n",
    "quantiles / 3600"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "7b36d5d4-ec27-4a59-b63f-515830da1ae9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.274951Z",
     "iopub.status.busy": "2022-12-07T15:27:46.274518Z",
     "iopub.status.idle": "2022-12-07T15:27:46.278079Z",
     "shell.execute_reply": "2022-12-07T15:27:46.277326Z"
    }
   },
   "outputs": [],
   "source": [
    "intervals = quantiles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "acc2bfd4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.281389Z",
     "iopub.status.busy": "2022-12-07T15:27:46.281127Z",
     "iopub.status.idle": "2022-12-07T15:27:46.287573Z",
     "shell.execute_reply": "2022-12-07T15:27:46.286822Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0          NaN\n",
       "0.1     0.057778\n",
       "0.2     0.239167\n",
       "0.3     0.377833\n",
       "0.4     0.313000\n",
       "0.5     0.608889\n",
       "0.6     0.764056\n",
       "0.7     7.578139\n",
       "0.8    10.252306\n",
       "0.9    48.948278\n",
       "Name: time_to_merge, dtype: float64"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# gaps between the buckets (i.e. size of each bucket) in hours\n",
    "(quantiles - quantiles.shift()) / 3600"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7111b3d8",
   "metadata": {},
   "source": [
    "If we want to frame our ML problem as a classification problem, we can use the above buckets to define the classes."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "21dc2dcf",
   "metadata": {},
   "source": [
    "**Body**\n",
    "\n",
    "Here, we will derive a feature called `body_size` which will simply be the number of words in the body of the PR. We can use this numerical value to describe the body of the PR."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "212f8a44",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.291772Z",
     "iopub.status.busy": "2022-12-07T15:27:46.291505Z",
     "iopub.status.idle": "2022-12-07T15:27:46.296418Z",
     "shell.execute_reply": "2022-12-07T15:27:46.295616Z"
    }
   },
   "outputs": [],
   "source": [
    "# derive body_size\n",
    "pr_df[\"body_size\"] = pr_df[\"body\"].fillna(\"\").apply(lambda x: len(x.split()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "c62d9484",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.299825Z",
     "iopub.status.busy": "2022-12-07T15:27:46.299559Z",
     "iopub.status.idle": "2022-12-07T15:27:46.305029Z",
     "shell.execute_reply": "2022-12-07T15:27:46.304292Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      41\n",
       "1       1\n",
       "2       1\n",
       "3       2\n",
       "4      53\n",
       "5       2\n",
       "6       3\n",
       "7      12\n",
       "8       6\n",
       "9      10\n",
       "10      2\n",
       "11     39\n",
       "12    124\n",
       "13      9\n",
       "14     15\n",
       "15     24\n",
       "16    153\n",
       "17      9\n",
       "18     26\n",
       "19     31\n",
       "20     19\n",
       "21      0\n",
       "22      7\n",
       "23     13\n",
       "24     47\n",
       "25     22\n",
       "26      2\n",
       "Name: body_size, dtype: int64"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pr_df[\"body_size\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9916fd09",
   "metadata": {},
   "source": [
    "**Size**\n",
    "\n",
    "We can attempt to convert the categorical size label generated by GitHub, which summarizes the size of the changes made to the code base, to a numerical value. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "cbc553ee-3e0b-45ed-83f6-02ae860bef85",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.309851Z",
     "iopub.status.busy": "2022-12-07T15:27:46.309583Z",
     "iopub.status.idle": "2022-12-07T15:27:46.314663Z",
     "shell.execute_reply": "2022-12-07T15:27:46.313906Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'L', 'M', 'S', 'XL', 'XS', 'XXL'}"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "set(pr_df['size'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "709b4cf9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.319084Z",
     "iopub.status.busy": "2022-12-07T15:27:46.318584Z",
     "iopub.status.idle": "2022-12-07T15:27:46.325466Z",
     "shell.execute_reply": "2022-12-07T15:27:46.324742Z"
    }
   },
   "outputs": [],
   "source": [
    "convert_size = {\"size\": {\"NaN\": 0, \"XS\": 0, \"S\": 1, \"M\": 2, \"L\": 3, \"XL\": 4, \"XXL\": 5}}\n",
    "pr_df = pr_df.replace(convert_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b803d48",
   "metadata": {},
   "source": [
    "**Title**\n",
    "\n",
    "Lets try to see if the PR title text can be used to extract any useful information regarding the PR."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "020f050a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.329023Z",
     "iopub.status.busy": "2022-12-07T15:27:46.328489Z",
     "iopub.status.idle": "2022-12-07T15:27:46.337096Z",
     "shell.execute_reply": "2022-12-07T15:27:46.336329Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0                          update inference output text\n",
       "1                               update inference yaml  \n",
       "2           testing  on pr  functionality  do not merge\n",
       "3                       add  with  mode  functionality \n",
       "4                      fix how inference was callingprs\n",
       "5                          fixed size condition for prs\n",
       "6                                  update entrypoint sh\n",
       "7                                  update entrypoint sh\n",
       "8                                  update entrypoint sh\n",
       "9                                       test train mode\n",
       "10    add training and inference modes in entrypoint sh\n",
       "11                  minor updates  fix number api calls\n",
       "12                         github api rate optimization\n",
       "13                                update inference yaml\n",
       "14           update inference yaml to use action inputs\n",
       "15                 added github api rate limit handling\n",
       "16    changed data download from srcopsmetrics to py...\n",
       "17                                    update action yml\n",
       "18                                  cleanup action test\n",
       "19                                         basic action\n",
       "20                        test pr trigger from upstream\n",
       "21                           sample pr to test workflow\n",
       "22                                update inference yaml\n",
       "23                          pr inference comment test 1\n",
       "24                             add inference for new pr\n",
       "25    clean feature engineering and choose best mode...\n",
       "26               add inference notebook to the workflow\n",
       "Name: title, dtype: object"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# first lets preprocess the text available in the titles\n",
    "\n",
    "# convert to lowercase\n",
    "preproc_titles = pr_df[\"title\"].str.lower()\n",
    "\n",
    "# remove punctuations and symbols like : ; , # ( ) [ ] etc\n",
    "preproc_titles = preproc_titles.str.replace(r'[`#-.?!,:;\\/()\\[\\]\"\\']', \" \", regex=True)\n",
    "\n",
    "# remove hash-like strings i.e. 25+ char long strings containing 0-9 or a-f\n",
    "preproc_titles = preproc_titles.str.replace(r\"[0-9a-f]{25,}\", \" \", regex=True)\n",
    "\n",
    "preproc_titles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "2ecc397e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.341739Z",
     "iopub.status.busy": "2022-12-07T15:27:46.341080Z",
     "iopub.status.idle": "2022-12-07T15:27:46.347587Z",
     "shell.execute_reply": "2022-12-07T15:27:46.346799Z"
    }
   },
   "outputs": [],
   "source": [
    "# what are the most commonly appearing words in the titles\n",
    "\n",
    "# combin all titles and split into words\n",
    "words = preproc_titles.str.cat(sep=\" \").split()\n",
    "\n",
    "# remove stopwords and numbers (e.g. bugzilla ids)\n",
    "words = [w for w in words if w not in set(STOPWORDS) and not w.isnumeric()]\n",
    "\n",
    "# word frequencies\n",
    "unique_words, counts = np.unique(words, return_counts=True)\n",
    "vc = pd.Series(counts, index=unique_words).sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c52f211c",
   "metadata": {},
   "source": [
    "From the above graph, we can see that titles often contain some keywords that can hint towards the type of changes being made in the PR. We will create a feature which is a vector indicating how many times each word appeared in the title. This way, even if a less prominent word is more correlated with time_to_merge, we will be able to capture that relationship."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "65279268",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.351425Z",
     "iopub.status.busy": "2022-12-07T15:27:46.350792Z",
     "iopub.status.idle": "2022-12-07T15:27:46.389971Z",
     "shell.execute_reply": "2022-12-07T15:27:46.389333Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 59/59 [00:00<00:00, 1898.27it/s]\n"
     ]
    }
   ],
   "source": [
    "# add word count columns\n",
    "for word in tqdm(unique_words):\n",
    "    pr_df[f\"title_wordcount_{word}\"] = preproc_titles.apply(\n",
    "        lambda x: x.split().count(word)\n",
    "    )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "41cf9f16",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.393560Z",
     "iopub.status.busy": "2022-12-07T15:27:46.392816Z",
     "iopub.status.idle": "2022-12-07T15:27:46.400702Z",
     "shell.execute_reply": "2022-12-07T15:27:46.400095Z"
    }
   },
   "outputs": [],
   "source": [
    "# collapse the high dim vector into one column\n",
    "wordcount_columns = [c for c in pr_df.columns if \"wordcount\" in c]\n",
    "pr_df[\"title_word_counts_vec\"] = pr_df[wordcount_columns].apply(\n",
    "    lambda x: x.tolist(), axis=1\n",
    ")\n",
    "\n",
    "# drop the individual wordcount columns\n",
    "pr_df = pr_df.drop(columns=wordcount_columns)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "06f46edc",
   "metadata": {},
   "source": [
    "**Created At**\n",
    "\n",
    "This column can give us information on when the PR was created. Using the `created_at` field, we can derive other features like `time_of_day`, `day_of_week`, `day_of_month`, `month_in_year` each of which can be categorical features that can be useful to our analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "b6db783b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.404142Z",
     "iopub.status.busy": "2022-12-07T15:27:46.403618Z",
     "iopub.status.idle": "2022-12-07T15:27:46.412600Z",
     "shell.execute_reply": "2022-12-07T15:27:46.411977Z"
    }
   },
   "outputs": [],
   "source": [
    "# convert to datetime object\n",
    "pr_df[\"created_at_datetime\"] = (\n",
    "    pr_df[\"created_at\"].fillna(\"\").apply(lambda x: datetime.datetime.fromtimestamp(x))\n",
    ")\n",
    "# get day as categorical variable\n",
    "pr_df[\"created_at_day\"] = pr_df[\"created_at_datetime\"].apply(lambda x: x.day)\n",
    "\n",
    "# get month as categorical variable\n",
    "pr_df[\"created_at_month\"] = pr_df[\"created_at_datetime\"].apply(lambda x: x.month)\n",
    "\n",
    "# get weekday as categorical variable\n",
    "pr_df[\"created_at_weekday\"] = pr_df[\"created_at_datetime\"].apply(lambda x: x.weekday())\n",
    "\n",
    "# get hour of day as categorical variable\n",
    "pr_df[\"created_at_hour\"] = pr_df[\"created_at_datetime\"].apply(lambda x: x.hour)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "f6cb1704",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.416072Z",
     "iopub.status.busy": "2022-12-07T15:27:46.415527Z",
     "iopub.status.idle": "2022-12-07T15:27:46.434992Z",
     "shell.execute_reply": "2022-12-07T15:27:46.434152Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>body</th>\n",
       "      <th>size</th>\n",
       "      <th>created_by</th>\n",
       "      <th>created_at</th>\n",
       "      <th>closed_at</th>\n",
       "      <th>closed_by</th>\n",
       "      <th>merged_at</th>\n",
       "      <th>merged_by</th>\n",
       "      <th>commits_number</th>\n",
       "      <th>...</th>\n",
       "      <th>first_review_at</th>\n",
       "      <th>first_approve_at</th>\n",
       "      <th>time_to_merge</th>\n",
       "      <th>body_size</th>\n",
       "      <th>title_word_counts_vec</th>\n",
       "      <th>created_at_datetime</th>\n",
       "      <th>created_at_day</th>\n",
       "      <th>created_at_month</th>\n",
       "      <th>created_at_weekday</th>\n",
       "      <th>created_at_hour</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>update inference output text</td>\n",
       "      <td>Updated `03_model_training.ipynb` to create a ...</td>\n",
       "      <td>5</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1670362761</td>\n",
       "      <td>1670426174</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1.670426e+09</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1.670363e+09</td>\n",
       "      <td>1.670367e+09</td>\n",
       "      <td>63413.0</td>\n",
       "      <td>41</td>\n",
       "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>2022-12-06 21:39:21</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Update inference.yaml.</td>\n",
       "      <td>test</td>\n",
       "      <td>1</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1670273067</td>\n",
       "      <td>1670276461</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1.670276e+09</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>1.670276e+09</td>\n",
       "      <td>1.670276e+09</td>\n",
       "      <td>3394.0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>2022-12-05 20:44:27</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 26 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                          title  \\\n",
       "0  update inference output text   \n",
       "1       Update inference.yaml.    \n",
       "\n",
       "                                                body  size       created_by  \\\n",
       "0  Updated `03_model_training.ipynb` to create a ...     5  MichaelClifford   \n",
       "1                                              test      1  aakankshaduggal   \n",
       "\n",
       "   created_at   closed_at        closed_by     merged_at        merged_by  \\\n",
       "0  1670362761  1670426174  MichaelClifford  1.670426e+09  MichaelClifford   \n",
       "1  1670273067  1670276461  MichaelClifford  1.670276e+09  MichaelClifford   \n",
       "\n",
       "   commits_number  ...  first_review_at first_approve_at time_to_merge  \\\n",
       "0               1  ...     1.670363e+09     1.670367e+09       63413.0   \n",
       "1               4  ...     1.670276e+09     1.670276e+09        3394.0   \n",
       "\n",
       "  body_size                              title_word_counts_vec  \\\n",
       "0        41  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...   \n",
       "1         1  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...   \n",
       "\n",
       "  created_at_datetime  created_at_day  created_at_month  created_at_weekday  \\\n",
       "0 2022-12-06 21:39:21               6                12                   1   \n",
       "1 2022-12-05 20:44:27               5                12                   0   \n",
       "\n",
       "   created_at_hour  \n",
       "0               21  \n",
       "1               20  \n",
       "\n",
       "[2 rows x 26 columns]"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pr_df.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "772ea2a9",
   "metadata": {},
   "source": [
    "**Changed Files Number**  \n",
    "Changed files is a numberical variable which captures the numer of files that were modified in the PR."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "b5687644",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.438997Z",
     "iopub.status.busy": "2022-12-07T15:27:46.438439Z",
     "iopub.status.idle": "2022-12-07T15:27:46.445586Z",
     "shell.execute_reply": "2022-12-07T15:27:46.444828Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1     11\n",
       "2      9\n",
       "4      4\n",
       "7      1\n",
       "5      1\n",
       "16     1\n",
       "Name: changed_files_number, dtype: int64"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# convert changed_files_number to int\n",
    "pr_df[\"changed_files_number\"] = pr_df[\"changed_files_number\"].astype(\"int\")\n",
    "\n",
    "pr_df.changed_files_number.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2ec4a803",
   "metadata": {},
   "source": [
    "`changed_files_number` has more variability than the `commits_number`. We can incorporate this as a numerical feature."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "360ea47c",
   "metadata": {},
   "source": [
    "**Changed_files**\n",
    "\n",
    "`changed_files` is a list of the paths and filenames for every file changed by this PR. From `changed_files`, we know, for each PR, which file types were changed which can be encoded as a frequency vector. Also we know which directory the change is made in, so that can be encoded as a categorical variable."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "ea080575",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.450186Z",
     "iopub.status.busy": "2022-12-07T15:27:46.449457Z",
     "iopub.status.idle": "2022-12-07T15:27:46.458800Z",
     "shell.execute_reply": "2022-12-07T15:27:46.458035Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0     [03_model_training.ipynb, 04_model_inference.i...\n",
       "1        [.github/workflows/inference.yaml, action.yml]\n",
       "2                                       [test_file.txt]\n",
       "3        [.github/workflows/inference.yaml, action.yml]\n",
       "4                            [04_model_inference.ipynb]\n",
       "5                                       [process_pr.py]\n",
       "6                                       [entrypoint.sh]\n",
       "7                                       [entrypoint.sh]\n",
       "8                                       [entrypoint.sh]\n",
       "9     [.github/workflows/inference.yaml, entrypoint.sh]\n",
       "10    [.github/workflows/inference.yaml, entrypoint.sh]\n",
       "11    [01_data_collection.py, 02_notebook_executed.i...\n",
       "12    [.github/workflows/run-in-container.yaml, 01_d...\n",
       "13       [.github/workflows/inference.yaml, action.yml]\n",
       "14                   [.github/workflows/inference.yaml]\n",
       "15    [01_data_collection.py, 02_notebook_executed.i...\n",
       "16    [.github/workflows/run-in-container.yaml, 01_d...\n",
       "17                                         [action.yml]\n",
       "18    [.github/workflows/inference.yaml, Dockerfile,...\n",
       "19    [Containerfile, action.yml, entrypoint.sh, tes...\n",
       "20                   [.github/workflows/inference.yaml]\n",
       "21                                          [README.md]\n",
       "22                   [.github/workflows/inference.yaml]\n",
       "23    [.github/workflows/inference.yaml, 04_model_in...\n",
       "24         [04_model_inference.ipynb, requirements.txt]\n",
       "25    [02_feature_engineering.ipynb, 03_model_traini...\n",
       "26    [.github/workflows/install-from-config.yaml, ....\n",
       "Name: changed_files, dtype: object"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pr_df.changed_files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "ba70b93f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.462623Z",
     "iopub.status.busy": "2022-12-07T15:27:46.462010Z",
     "iopub.status.idle": "2022-12-07T15:27:46.467644Z",
     "shell.execute_reply": "2022-12-07T15:27:46.466871Z"
    }
   },
   "outputs": [],
   "source": [
    "def filetype(filepath):\n",
    "\n",
    "    # if standard file extension, return file extension\n",
    "    if pathlib.Path(filepath).suffix:\n",
    "        return pathlib.Path(filepath).suffix\n",
    "\n",
    "    else:\n",
    "        # if single file at the root of repo\n",
    "        if \"/\" not in filepath:\n",
    "            # if a file like README/Dockerfile etc\n",
    "            if \".\" not in filepath:\n",
    "                return filepath\n",
    "        # if file not at root\n",
    "        else:\n",
    "            # get file name\n",
    "            file = filepath.rsplit(\"/\", 1)[1]\n",
    "            if \".\" not in file:\n",
    "                return file\n",
    "            elif \".\" in file:\n",
    "                return \".\" + file.rsplit(\".\", 1)[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "2e00c415",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.471317Z",
     "iopub.status.busy": "2022-12-07T15:27:46.470705Z",
     "iopub.status.idle": "2022-12-07T15:27:46.474412Z",
     "shell.execute_reply": "2022-12-07T15:27:46.473633Z"
    }
   },
   "outputs": [],
   "source": [
    "# get list of all file types that were modified across all openshift origin PRs\n",
    "changed_files = list(pr_df[\"changed_files\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "36a1dcda",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.478099Z",
     "iopub.status.busy": "2022-12-07T15:27:46.477392Z",
     "iopub.status.idle": "2022-12-07T15:27:46.482040Z",
     "shell.execute_reply": "2022-12-07T15:27:46.481323Z"
    }
   },
   "outputs": [],
   "source": [
    "all_changed_files = [filetype(filepath) for pr in changed_files for filepath in pr]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "d7de1e91",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.485202Z",
     "iopub.status.busy": "2022-12-07T15:27:46.484763Z",
     "iopub.status.idle": "2022-12-07T15:27:46.488329Z",
     "shell.execute_reply": "2022-12-07T15:27:46.487559Z"
    }
   },
   "outputs": [],
   "source": [
    "count = Counter(all_changed_files)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "17e5fd80",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.491862Z",
     "iopub.status.busy": "2022-12-07T15:27:46.491265Z",
     "iopub.status.idle": "2022-12-07T15:27:46.495104Z",
     "shell.execute_reply": "2022-12-07T15:27:46.494342Z"
    }
   },
   "outputs": [],
   "source": [
    "# save 100 most commonly occuring file extensions across PRs\n",
    "top_fileextensions_freq = count.most_common(100)\n",
    "top_fileextensions = [ext[0] for ext in top_fileextensions_freq]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "25a14da8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.498635Z",
     "iopub.status.busy": "2022-12-07T15:27:46.498019Z",
     "iopub.status.idle": "2022-12-07T15:27:46.503073Z",
     "shell.execute_reply": "2022-12-07T15:27:46.502319Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['.ipynb',\n",
       " '.yaml',\n",
       " '.py',\n",
       " '.sh',\n",
       " '.yml',\n",
       " '.txt',\n",
       " '.md',\n",
       " 'LICENSE',\n",
       " 'Dockerfile',\n",
       " 'Containerfile']"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_fileextensions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "ecd11700",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.507145Z",
     "iopub.status.busy": "2022-12-07T15:27:46.506735Z",
     "iopub.status.idle": "2022-12-07T15:27:46.511782Z",
     "shell.execute_reply": "2022-12-07T15:27:46.511032Z"
    }
   },
   "outputs": [],
   "source": [
    "def file_type_freq(list_of_filepaths):\n",
    "    \"\"\"\n",
    "    For a given PR's list of changed files,\n",
    "    return a changed files frequency vector\n",
    "    ( a vector which holds the frequency of\n",
    "    changed files of the type  for every\n",
    "    file type in top_fileextensions )\n",
    "    \"\"\"\n",
    "\n",
    "    file_extensions = [filetype(f) for f in list_of_filepaths]\n",
    "    ext_dict = {key: 0 for key in top_fileextensions}\n",
    "    for f in file_extensions:\n",
    "        if f in ext_dict:\n",
    "            ext_dict[f] += 1\n",
    "\n",
    "    return list(ext_dict.values())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "cc0b7906",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.515073Z",
     "iopub.status.busy": "2022-12-07T15:27:46.514817Z",
     "iopub.status.idle": "2022-12-07T15:27:46.520004Z",
     "shell.execute_reply": "2022-12-07T15:27:46.519227Z"
    }
   },
   "outputs": [],
   "source": [
    "pr_df[\"changed_file_type_vec\"] = pr_df[\"changed_files\"].apply(\n",
    "    lambda x: file_type_freq(x)\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "4fe7bb1d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.523311Z",
     "iopub.status.busy": "2022-12-07T15:27:46.523052Z",
     "iopub.status.idle": "2022-12-07T15:27:46.543058Z",
     "shell.execute_reply": "2022-12-07T15:27:46.542240Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>body</th>\n",
       "      <th>size</th>\n",
       "      <th>created_by</th>\n",
       "      <th>created_at</th>\n",
       "      <th>closed_at</th>\n",
       "      <th>closed_by</th>\n",
       "      <th>merged_at</th>\n",
       "      <th>merged_by</th>\n",
       "      <th>commits_number</th>\n",
       "      <th>...</th>\n",
       "      <th>first_approve_at</th>\n",
       "      <th>time_to_merge</th>\n",
       "      <th>body_size</th>\n",
       "      <th>title_word_counts_vec</th>\n",
       "      <th>created_at_datetime</th>\n",
       "      <th>created_at_day</th>\n",
       "      <th>created_at_month</th>\n",
       "      <th>created_at_weekday</th>\n",
       "      <th>created_at_hour</th>\n",
       "      <th>changed_file_type_vec</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>update inference output text</td>\n",
       "      <td>Updated `03_model_training.ipynb` to create a ...</td>\n",
       "      <td>5</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1670362761</td>\n",
       "      <td>1670426174</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1.670426e+09</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>1.670367e+09</td>\n",
       "      <td>63413.0</td>\n",
       "      <td>41</td>\n",
       "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>2022-12-06 21:39:21</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>21</td>\n",
       "      <td>[2, 0, 0, 0, 0, 0, 0, 0, 0, 0]</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Update inference.yaml.</td>\n",
       "      <td>test</td>\n",
       "      <td>1</td>\n",
       "      <td>aakankshaduggal</td>\n",
       "      <td>1670273067</td>\n",
       "      <td>1670276461</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>1.670276e+09</td>\n",
       "      <td>MichaelClifford</td>\n",
       "      <td>4</td>\n",
       "      <td>...</td>\n",
       "      <td>1.670276e+09</td>\n",
       "      <td>3394.0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>2022-12-05 20:44:27</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>[0, 1, 0, 0, 1, 0, 0, 0, 0, 0]</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>2 rows × 27 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                          title  \\\n",
       "0  update inference output text   \n",
       "1       Update inference.yaml.    \n",
       "\n",
       "                                                body  size       created_by  \\\n",
       "0  Updated `03_model_training.ipynb` to create a ...     5  MichaelClifford   \n",
       "1                                              test      1  aakankshaduggal   \n",
       "\n",
       "   created_at   closed_at        closed_by     merged_at        merged_by  \\\n",
       "0  1670362761  1670426174  MichaelClifford  1.670426e+09  MichaelClifford   \n",
       "1  1670273067  1670276461  MichaelClifford  1.670276e+09  MichaelClifford   \n",
       "\n",
       "   commits_number  ...  first_approve_at time_to_merge body_size  \\\n",
       "0               1  ...      1.670367e+09       63413.0        41   \n",
       "1               4  ...      1.670276e+09        3394.0         1   \n",
       "\n",
       "                               title_word_counts_vec created_at_datetime  \\\n",
       "0  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ... 2022-12-06 21:39:21   \n",
       "1  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ... 2022-12-05 20:44:27   \n",
       "\n",
       "  created_at_day  created_at_month  created_at_weekday  created_at_hour  \\\n",
       "0              6                12                   1               21   \n",
       "1              5                12                   0               20   \n",
       "\n",
       "            changed_file_type_vec  \n",
       "0  [2, 0, 0, 0, 0, 0, 0, 0, 0, 0]  \n",
       "1  [0, 1, 0, 0, 1, 0, 0, 0, 0, 0]  \n",
       "\n",
       "[2 rows x 27 columns]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pr_df.head(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8102f95e-6175-4b46-99c1-66175f20325c",
   "metadata": {},
   "source": [
    "## Create PR trends dashboard"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e502b7f-49c4-4024-aebc-3c9720a94209",
   "metadata": {},
   "source": [
    "## Infer dataframe columns as Features"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "70443afd",
   "metadata": {},
   "source": [
    "We can see above that some PRs have 0 for all directory fields since the directory structure of the repo has changed over time and we are interested in only the latest directory structure."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "64e7a822",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.547819Z",
     "iopub.status.busy": "2022-12-07T15:27:46.547294Z",
     "iopub.status.idle": "2022-12-07T15:27:46.552811Z",
     "shell.execute_reply": "2022-12-07T15:27:46.552018Z"
    }
   },
   "outputs": [],
   "source": [
    "# subset DataFrame to use as features\n",
    "feature_df = pr_df[\n",
    "    [\n",
    "        \"time_to_merge\",\n",
    "        \"changed_files_number\",\n",
    "        \"body_size\",\n",
    "        \"size\",\n",
    "        \"title_word_counts_vec\",\n",
    "        \"created_at_day\",\n",
    "        \"created_at_month\",\n",
    "        \"created_at_weekday\",\n",
    "        \"created_at_hour\",\n",
    "        \"commits_number\",\n",
    "        \"changed_file_type_vec\",\n",
    "    ]\n",
    "]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0b08a241",
   "metadata": {},
   "source": [
    "### Add PR TTM Class Column \n",
    "\n",
    "Lets split up the time to merge values into 10 discrete bins (as discussed above) to use as our prediction classes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "b1a5adc0-e8e2-4159-af36-a6fdc938de17",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.556572Z",
     "iopub.status.busy": "2022-12-07T15:27:46.556125Z",
     "iopub.status.idle": "2022-12-07T15:27:46.564686Z",
     "shell.execute_reply": "2022-12-07T15:27:46.564006Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0        93.0\n",
       "0.1       301.0\n",
       "0.2      1162.0\n",
       "0.3      2522.2\n",
       "0.4      3649.0\n",
       "0.5      5841.0\n",
       "0.6      8591.6\n",
       "0.7     35872.9\n",
       "0.8     72781.2\n",
       "0.9    248995.0\n",
       "Name: time_to_merge, dtype: float64"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intervals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "ec16152d-0fb3-4775-913a-9e476382d713",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.568701Z",
     "iopub.status.busy": "2022-12-07T15:27:46.568083Z",
     "iopub.status.idle": "2022-12-07T15:27:46.793609Z",
     "shell.execute_reply": "2022-12-07T15:27:46.792846Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_143/2436122750.py:9: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  feature_df[\"ttm_class\"] = (feature_df[\"time_to_merge\"]).apply(get_ttm_class)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>time_to_merge</th>\n",
       "      <th>changed_files_number</th>\n",
       "      <th>body_size</th>\n",
       "      <th>size</th>\n",
       "      <th>title_word_counts_vec</th>\n",
       "      <th>created_at_day</th>\n",
       "      <th>created_at_month</th>\n",
       "      <th>created_at_weekday</th>\n",
       "      <th>created_at_hour</th>\n",
       "      <th>commits_number</th>\n",
       "      <th>changed_file_type_vec</th>\n",
       "      <th>ttm_class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>63413.0</td>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "      <td>5</td>\n",
       "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>21</td>\n",
       "      <td>1</td>\n",
       "      <td>[2, 0, 0, 0, 0, 0, 0, 0, 0, 0]</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>3394.0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>4</td>\n",
       "      <td>[0, 1, 0, 0, 1, 0, 0, 0, 0, 0]</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>NaN</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>1</td>\n",
       "      <td>[0, 0, 0, 0, 0, 1, 0, 0, 0, 0]</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NaN</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>[0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>3</td>\n",
       "      <td>[0, 1, 0, 0, 1, 0, 0, 0, 0, 0]</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>390.0</td>\n",
       "      <td>1</td>\n",
       "      <td>53</td>\n",
       "      <td>3</td>\n",
       "      <td>[0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, ...</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>3</td>\n",
       "      <td>15</td>\n",
       "      <td>1</td>\n",
       "      <td>[1, 0, 0, 0, 0, 0, 0, 0, 0, 0]</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   time_to_merge  changed_files_number  body_size  size  \\\n",
       "0        63413.0                     2         41     5   \n",
       "1         3394.0                     2          1     1   \n",
       "2            NaN                     1          1     0   \n",
       "3            NaN                     2          2     1   \n",
       "4          390.0                     1         53     3   \n",
       "\n",
       "                               title_word_counts_vec  created_at_day  \\\n",
       "0  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...               6   \n",
       "1  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...               5   \n",
       "2  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...               5   \n",
       "3  [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...               5   \n",
       "4  [0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, ...               1   \n",
       "\n",
       "   created_at_month  created_at_weekday  created_at_hour  commits_number  \\\n",
       "0                12                   1               21               1   \n",
       "1                12                   0               20               4   \n",
       "2                12                   0               20               1   \n",
       "3                12                   0               17               3   \n",
       "4                12                   3               15               1   \n",
       "\n",
       "            changed_file_type_vec  ttm_class  \n",
       "0  [2, 0, 0, 0, 0, 0, 0, 0, 0, 0]          7  \n",
       "1  [0, 1, 0, 0, 1, 0, 0, 0, 0, 0]          3  \n",
       "2  [0, 0, 0, 0, 0, 1, 0, 0, 0, 0]          9  \n",
       "3  [0, 1, 0, 0, 1, 0, 0, 0, 0, 0]          9  \n",
       "4  [1, 0, 0, 0, 0, 0, 0, 0, 0, 0]          1  "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_ttm_class(ttm):\n",
    "\n",
    "    for rownum, (indx, val) in enumerate(intervals.iteritems()):\n",
    "        if ttm < val:\n",
    "            return rownum - 1\n",
    "    return 9\n",
    "\n",
    "\n",
    "feature_df[\"ttm_class\"] = (feature_df[\"time_to_merge\"]).apply(get_ttm_class)\n",
    "feature_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "04749f45",
   "metadata": {},
   "source": [
    "## Evaluating Feature Importances\n",
    "\n",
    "### Categorical Features\n",
    "\n",
    "First, lets try to determine what categorical featuers are powerful influencers for time to merge of a PR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "e353c92b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.797688Z",
     "iopub.status.busy": "2022-12-07T15:27:46.797236Z",
     "iopub.status.idle": "2022-12-07T15:27:46.802982Z",
     "shell.execute_reply": "2022-12-07T15:27:46.802276Z"
    }
   },
   "outputs": [],
   "source": [
    "# feature selection helper function\n",
    "def select_features(x, y, func):\n",
    "    fs = SelectKBest(score_func=func, k=\"all\")\n",
    "    fs.fit(x, y)\n",
    "    x_fs = fs.transform(x)\n",
    "    return x_fs, fs\n",
    "\n",
    "\n",
    "selected_features = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "986b3424-0401-4c7a-8f28-a860b87e14ef",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.806870Z",
     "iopub.status.busy": "2022-12-07T15:27:46.806147Z",
     "iopub.status.idle": "2022-12-07T15:27:46.810709Z",
     "shell.execute_reply": "2022-12-07T15:27:46.810048Z"
    }
   },
   "outputs": [],
   "source": [
    "cat_features = [\n",
    "    \"size\",\n",
    "    \"created_at_day\",\n",
    "    \"created_at_month\",\n",
    "    \"created_at_weekday\",\n",
    "    \"created_at_hour\",\n",
    "    \"ttm_class\",\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "3e1d24a4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.814604Z",
     "iopub.status.busy": "2022-12-07T15:27:46.814024Z",
     "iopub.status.idle": "2022-12-07T15:27:46.825465Z",
     "shell.execute_reply": "2022-12-07T15:27:46.824688Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>size</th>\n",
       "      <th>created_at_day</th>\n",
       "      <th>created_at_month</th>\n",
       "      <th>created_at_weekday</th>\n",
       "      <th>created_at_hour</th>\n",
       "      <th>ttm_class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>21</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>3</td>\n",
       "      <td>15</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   size  created_at_day  created_at_month  created_at_weekday  \\\n",
       "0     5               6                12                   1   \n",
       "1     1               5                12                   0   \n",
       "2     0               5                12                   0   \n",
       "3     1               5                12                   0   \n",
       "4     3               1                12                   3   \n",
       "\n",
       "   created_at_hour  ttm_class  \n",
       "0               21          7  \n",
       "1               20          3  \n",
       "2               20          9  \n",
       "3               17          9  \n",
       "4               15          1  "
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cat_df = feature_df[cat_features]\n",
    "\n",
    "cat_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "278e3073",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.829082Z",
     "iopub.status.busy": "2022-12-07T15:27:46.828811Z",
     "iopub.status.idle": "2022-12-07T15:27:46.834545Z",
     "shell.execute_reply": "2022-12-07T15:27:46.833831Z"
    }
   },
   "outputs": [],
   "source": [
    "# split into input (x) and output (y) variables\n",
    "X_inp = cat_df.drop(columns=[\"ttm_class\"])\n",
    "y = cat_df[\"ttm_class\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "2df4fb8d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:46.838107Z",
     "iopub.status.busy": "2022-12-07T15:27:46.837494Z",
     "iopub.status.idle": "2022-12-07T15:27:47.047011Z",
     "shell.execute_reply": "2022-12-07T15:27:47.045990Z"
    }
   },
   "outputs": [],
   "source": [
    "# rank using mutual_info_classif\n",
    "X_fs, fs = select_features(X_inp, y, mutual_info_classif)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "418becb1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.051453Z",
     "iopub.status.busy": "2022-12-07T15:27:47.050846Z",
     "iopub.status.idle": "2022-12-07T15:27:47.057460Z",
     "shell.execute_reply": "2022-12-07T15:27:47.056654Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature size: 0.294949\n",
      "Feature created_at_day: 0.286521\n",
      "Feature created_at_month: 0.562814\n",
      "Feature created_at_weekday: 0.155237\n",
      "Feature created_at_hour: 0.000000\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(fs.scores_)):\n",
    "    print(\"Feature %s: %f\" % (X_inp.columns[i], fs.scores_[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "b08ea157",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.060915Z",
     "iopub.status.busy": "2022-12-07T15:27:47.060459Z",
     "iopub.status.idle": "2022-12-07T15:27:47.220336Z",
     "shell.execute_reply": "2022-12-07T15:27:47.219594Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize scores of all features\n",
    "sns.barplot(x=X_inp.columns, y=fs.scores_)\n",
    "plt.title(\"Feature importance scores for Mutual Information Statistic\")\n",
    "plt.xticks(rotation=60)\n",
    "plt.xlabel(\"Features\")\n",
    "plt.ylabel(\"Scores\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "176c1d98",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.224030Z",
     "iopub.status.busy": "2022-12-07T15:27:47.223461Z",
     "iopub.status.idle": "2022-12-07T15:27:47.227397Z",
     "shell.execute_reply": "2022-12-07T15:27:47.226780Z"
    }
   },
   "outputs": [],
   "source": [
    "# Lets get a list of the top 10 feature indices based on Mutual Info Stat feature scores\n",
    "top_mis = fs.scores_.argsort()[-10:][::-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "c33dc4f2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.230698Z",
     "iopub.status.busy": "2022-12-07T15:27:47.230182Z",
     "iopub.status.idle": "2022-12-07T15:27:47.237941Z",
     "shell.execute_reply": "2022-12-07T15:27:47.237323Z"
    }
   },
   "outputs": [],
   "source": [
    "# rank using chi_sq\n",
    "X_fs, fs = select_features(X_inp, y, chi2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "313c6906",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.241327Z",
     "iopub.status.busy": "2022-12-07T15:27:47.240800Z",
     "iopub.status.idle": "2022-12-07T15:27:47.245436Z",
     "shell.execute_reply": "2022-12-07T15:27:47.244840Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Feature size: 27.967213\n",
      "Feature created_at_day: 66.087500\n",
      "Feature created_at_month: 0.091362\n",
      "Feature created_at_weekday: 5.304688\n",
      "Feature created_at_hour: 3.253074\n"
     ]
    }
   ],
   "source": [
    "for i in range(len(fs.scores_)):\n",
    "    print(\"Feature %s: %f\" % (X_inp.columns[i], fs.scores_[i]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "460daed6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.248996Z",
     "iopub.status.busy": "2022-12-07T15:27:47.248205Z",
     "iopub.status.idle": "2022-12-07T15:27:47.386265Z",
     "shell.execute_reply": "2022-12-07T15:27:47.385563Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize scores of all features\n",
    "sns.barplot(x=X_inp.columns, y=fs.scores_)\n",
    "plt.title(\"Feature importance scores for Chi Squared\")\n",
    "plt.xticks(rotation=60)\n",
    "plt.xlabel(\"Features\")\n",
    "plt.ylabel(\"Scores\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "9874b4b4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.390212Z",
     "iopub.status.busy": "2022-12-07T15:27:47.389379Z",
     "iopub.status.idle": "2022-12-07T15:27:47.393488Z",
     "shell.execute_reply": "2022-12-07T15:27:47.392845Z"
    }
   },
   "outputs": [],
   "source": [
    "# Lets get a list of the top 10 feature indices based on Chi Sq feature scores\n",
    "top_chi2 = fs.scores_.argsort()[-10:][::-1]"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5861b291",
   "metadata": {},
   "source": [
    "Let's get a Union of the top 10 features derived from features based on both scores of both Mutual Info Statistic and Chi Squared"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "fae59295",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.397319Z",
     "iopub.status.busy": "2022-12-07T15:27:47.396507Z",
     "iopub.status.idle": "2022-12-07T15:27:47.402746Z",
     "shell.execute_reply": "2022-12-07T15:27:47.402128Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{0, 1, 2, 3, 4}"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "union = set(top_mis).union(set(top_chi2))\n",
    "print(len(union))\n",
    "union"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "4778b695-9ec8-4d9f-a1dd-26d749b43a95",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.406329Z",
     "iopub.status.busy": "2022-12-07T15:27:47.405540Z",
     "iopub.status.idle": "2022-12-07T15:27:47.410956Z",
     "shell.execute_reply": "2022-12-07T15:27:47.410320Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['size',\n",
       " 'created_at_day',\n",
       " 'created_at_month',\n",
       " 'created_at_weekday',\n",
       " 'created_at_hour']"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_cat_features = [cat_features[i] for i in union]\n",
    "selected_cat_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "5b0ce173",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.414540Z",
     "iopub.status.busy": "2022-12-07T15:27:47.413741Z",
     "iopub.status.idle": "2022-12-07T15:27:47.417453Z",
     "shell.execute_reply": "2022-12-07T15:27:47.416819Z"
    }
   },
   "outputs": [],
   "source": [
    "# update list of features to keep for training\n",
    "selected_features += selected_cat_features"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "33f4387d",
   "metadata": {},
   "source": [
    "### Numerical Features\n",
    "\n",
    "Now, lets try to determine what numerical featuers are powerful influencers for time to merge of a PR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "29683916-4ef7-474a-80ac-11a5d210b6fa",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.421271Z",
     "iopub.status.busy": "2022-12-07T15:27:47.420457Z",
     "iopub.status.idle": "2022-12-07T15:27:47.424368Z",
     "shell.execute_reply": "2022-12-07T15:27:47.423715Z"
    }
   },
   "outputs": [],
   "source": [
    "numerical_features = [\n",
    "    \"changed_files_number\",\n",
    "    \"body_size\",\n",
    "    \"commits_number\",\n",
    "    \"ttm_class\",\n",
    "]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "b0d608dd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.427818Z",
     "iopub.status.busy": "2022-12-07T15:27:47.427270Z",
     "iopub.status.idle": "2022-12-07T15:27:47.436715Z",
     "shell.execute_reply": "2022-12-07T15:27:47.436063Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>changed_files_number</th>\n",
       "      <th>body_size</th>\n",
       "      <th>commits_number</th>\n",
       "      <th>ttm_class</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>53</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   changed_files_number  body_size  commits_number  ttm_class\n",
       "0                     2         41               1          7\n",
       "1                     2          1               4          3\n",
       "2                     1          1               1          9\n",
       "3                     2          2               3          9\n",
       "4                     1         53               1          1"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "numerical_df = feature_df[numerical_features]\n",
    "numerical_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "9ff2919e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.440178Z",
     "iopub.status.busy": "2022-12-07T15:27:47.439612Z",
     "iopub.status.idle": "2022-12-07T15:27:47.444394Z",
     "shell.execute_reply": "2022-12-07T15:27:47.443735Z"
    }
   },
   "outputs": [],
   "source": [
    "# split into input features and target\n",
    "X_inp = numerical_df.drop(columns=[\"ttm_class\"])\n",
    "y = numerical_df[\"ttm_class\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "cc733746",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.447821Z",
     "iopub.status.busy": "2022-12-07T15:27:47.447167Z",
     "iopub.status.idle": "2022-12-07T15:27:47.586918Z",
     "shell.execute_reply": "2022-12-07T15:27:47.586219Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create k best feature selector object based on mutual information\n",
    "numerical_feats_mi, k_best_selector_mi = select_features(\n",
    "    X_inp,\n",
    "    y,\n",
    "    partial(\n",
    "        mutual_info_classif,\n",
    "        discrete_features=False,\n",
    "        random_state=42,\n",
    "        n_neighbors=3,\n",
    "    ),\n",
    ")\n",
    "\n",
    "# visualize the MI scores of all features\n",
    "sns.barplot(x=X_inp.columns, y=k_best_selector_mi.scores_)\n",
    "plt.title(\"Feature importance based on Mutual Information\")\n",
    "plt.xticks(rotation=60)\n",
    "plt.xlabel(\"Features\")\n",
    "plt.ylabel(\"Scores\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "7bdf2192",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.590427Z",
     "iopub.status.busy": "2022-12-07T15:27:47.589963Z",
     "iopub.status.idle": "2022-12-07T15:27:47.767377Z",
     "shell.execute_reply": "2022-12-07T15:27:47.766691Z"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create k best feature selector object based on anova f statistic\n",
    "numerical_feats_f, k_best_selector_f = select_features(\n",
    "    X_inp,\n",
    "    y,\n",
    "    f_classif,\n",
    ")\n",
    "\n",
    "# visualize the f stat scores and pvalues of all features\n",
    "fig, axes = plt.subplots(2, 1, sharex=True)\n",
    "fig.suptitle(\"Feature importance based on ANOVA F stat\")\n",
    "sns.barplot(x=X_inp.columns, y=k_best_selector_f.scores_, ax=axes[0])\n",
    "sns.barplot(x=X_inp.columns, y=k_best_selector_f.pvalues_, ax=axes[1])\n",
    "plt.xticks(rotation=60)\n",
    "plt.xlabel(\"Features\")\n",
    "axes[0].set_ylabel(\"Scores\")\n",
    "axes[1].set_ylabel(\"p-values\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "6134ddb1-994f-42ee-8a68-7111540b6481",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.771179Z",
     "iopub.status.busy": "2022-12-07T15:27:47.770625Z",
     "iopub.status.idle": "2022-12-07T15:27:47.776685Z",
     "shell.execute_reply": "2022-12-07T15:27:47.776005Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{0, 1, 2}"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "top_mi = k_best_selector_mi.scores_.argsort()[-3:][::-1]\n",
    "top_anova = k_best_selector_f.scores_.argsort()[-3:][::-1]\n",
    "union = set(top_mi).union(set(top_anova))\n",
    "union"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "2d5a9368-bfaf-4348-b140-13e6d725f59c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.780190Z",
     "iopub.status.busy": "2022-12-07T15:27:47.779634Z",
     "iopub.status.idle": "2022-12-07T15:27:47.784589Z",
     "shell.execute_reply": "2022-12-07T15:27:47.784001Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['changed_files_number', 'body_size', 'commits_number']"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_num_features = [numerical_features[i] for i in union]\n",
    "selected_num_features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "id": "cc984a01",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.787967Z",
     "iopub.status.busy": "2022-12-07T15:27:47.787437Z",
     "iopub.status.idle": "2022-12-07T15:27:47.790845Z",
     "shell.execute_reply": "2022-12-07T15:27:47.790255Z"
    }
   },
   "outputs": [],
   "source": [
    "# update list of features to keep for training\n",
    "selected_features += selected_num_features"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "11a86b36",
   "metadata": {},
   "source": [
    "### Count Vector Type Features\n",
    "\n",
    "Finally, lets determine the most important features from the count vector type columns."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "id": "f07d0334",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.794447Z",
     "iopub.status.busy": "2022-12-07T15:27:47.793948Z",
     "iopub.status.idle": "2022-12-07T15:27:47.799418Z",
     "shell.execute_reply": "2022-12-07T15:27:47.798807Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# explode the list column so that there is 1 column representing frequency of each file type\n",
    "filetype_df = pd.DataFrame(feature_df.changed_file_type_vec.to_list())\n",
    "filetype_df.index = feature_df.index\n",
    "filetype_df.columns = [f\"filetype_{f}\" for f in top_fileextensions]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "id": "35f93d4e",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.803034Z",
     "iopub.status.busy": "2022-12-07T15:27:47.802511Z",
     "iopub.status.idle": "2022-12-07T15:27:47.812727Z",
     "shell.execute_reply": "2022-12-07T15:27:47.812103Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "filetype_.txt             0.373649\n",
       "filetype_.yaml            0.363663\n",
       "filetype_.md              0.298142\n",
       "filetype_LICENSE          0.206725\n",
       "filetype_.ipynb           0.194111\n",
       "filetype_.py              0.173720\n",
       "filetype_Dockerfile      -0.041345\n",
       "filetype_Containerfile   -0.103362\n",
       "filetype_.yml            -0.197203\n",
       "filetype_.sh             -0.365257\n",
       "dtype: float64"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compute correlation with ttm\n",
    "corr = filetype_df.corrwith(feature_df[\"ttm_class\"])\n",
    "corr.sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "id": "d5ae2981",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.816287Z",
     "iopub.status.busy": "2022-12-07T15:27:47.815731Z",
     "iopub.status.idle": "2022-12-07T15:27:47.822317Z",
     "shell.execute_reply": "2022-12-07T15:27:47.821699Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "filetype_.ipynb           0.194111\n",
       "filetype_.yaml            0.363663\n",
       "filetype_.py              0.173720\n",
       "filetype_.sh             -0.365257\n",
       "filetype_.yml            -0.197203\n",
       "filetype_.txt             0.373649\n",
       "filetype_.md              0.298142\n",
       "filetype_LICENSE          0.206725\n",
       "filetype_Dockerfile      -0.041345\n",
       "filetype_Containerfile   -0.103362\n",
       "dtype: float64"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr[corr.abs() > 0.01]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "id": "9d5e19bd",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.825789Z",
     "iopub.status.busy": "2022-12-07T15:27:47.825287Z",
     "iopub.status.idle": "2022-12-07T15:27:47.830315Z",
     "shell.execute_reply": "2022-12-07T15:27:47.829719Z"
    }
   },
   "outputs": [],
   "source": [
    "# update list of features to keep for training\n",
    "# take everything that has more than 0.01 correlation magnitude\n",
    "filetype_keep_cols = corr[corr.abs() > 0.01].index.tolist()\n",
    "selected_features += filetype_keep_cols\n",
    "\n",
    "# only keep the relevant columns now\n",
    "filetype_df = filetype_df[filetype_keep_cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "id": "2eab2fb8-ee9e-4a4e-9d2e-7629f9c56cb9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.833644Z",
     "iopub.status.busy": "2022-12-07T15:27:47.833152Z",
     "iopub.status.idle": "2022-12-07T15:27:47.838013Z",
     "shell.execute_reply": "2022-12-07T15:27:47.837449Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['.ipynb',\n",
       " '.yaml',\n",
       " '.py',\n",
       " '.sh',\n",
       " '.yml',\n",
       " '.txt',\n",
       " '.md',\n",
       " 'LICENSE',\n",
       " 'Dockerfile',\n",
       " 'Containerfile']"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filetype_keep_cols = [i[9:] for i in filetype_keep_cols]\n",
    "filetype_keep_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "id": "2b2da9ab-de61-44f5-91f1-a897e8f7a962",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.841272Z",
     "iopub.status.busy": "2022-12-07T15:27:47.840783Z",
     "iopub.status.idle": "2022-12-07T15:27:47.844664Z",
     "shell.execute_reply": "2022-12-07T15:27:47.844085Z"
    }
   },
   "outputs": [],
   "source": [
    "filetype_keep_cols_df = pd.DataFrame(filetype_keep_cols, columns=[\"filetype\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "id": "773682ad-cda8-42ec-9275-dc611004d913",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.848054Z",
     "iopub.status.busy": "2022-12-07T15:27:47.847529Z",
     "iopub.status.idle": "2022-12-07T15:27:47.854368Z",
     "shell.execute_reply": "2022-12-07T15:27:47.853654Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>filetype</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>.ipynb</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>.yaml</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>.py</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>.sh</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>.yml</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>.txt</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>.md</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>LICENSE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Dockerfile</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Containerfile</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        filetype\n",
       "0         .ipynb\n",
       "1          .yaml\n",
       "2            .py\n",
       "3            .sh\n",
       "4           .yml\n",
       "5           .txt\n",
       "6            .md\n",
       "7        LICENSE\n",
       "8     Dockerfile\n",
       "9  Containerfile"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filetype_keep_cols_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "id": "7333e449-4e2a-42aa-b728-3f36d205ed1f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:47.857691Z",
     "iopub.status.busy": "2022-12-07T15:27:47.857193Z",
     "iopub.status.idle": "2022-12-07T15:27:48.219347Z",
     "shell.execute_reply": "2022-12-07T15:27:48.218493Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ResponseMetadata': {'RequestId': 'lbdt10aw-g1m31l-vny',\n",
       "  'HostId': 'lbdt10aw-g1m31l-vny',\n",
       "  'HTTPStatusCode': 200,\n",
       "  'HTTPHeaders': {'x-amz-request-id': 'lbdt10aw-g1m31l-vny',\n",
       "   'x-amz-id-2': 'lbdt10aw-g1m31l-vny',\n",
       "   'access-control-allow-origin': '*',\n",
       "   'access-control-allow-credentials': 'true',\n",
       "   'access-control-allow-methods': 'GET,POST,PUT,DELETE,OPTIONS',\n",
       "   'access-control-allow-headers': 'Content-Type,Content-MD5,Authorization,X-Amz-User-Agent,X-Amz-Date,ETag,X-Amz-Content-Sha256',\n",
       "   'access-control-expose-headers': 'ETag,X-Amz-Version-Id',\n",
       "   'etag': '\"13c872c2169c9f1e43d1da390ef4ad2e\"',\n",
       "   'date': 'Wed, 07 Dec 2022 15:27:48 GMT',\n",
       "   'keep-alive': 'timeout=5',\n",
       "   'content-length': '0',\n",
       "   'set-cookie': '1a4aa612fe797ac8466d7ee00e5520d5=b9a4e3efe932781180a9a8945a639f83; path=/; HttpOnly; Secure; SameSite=None'},\n",
       "  'RetryAttempts': 0},\n",
       " 'ETag': '\"13c872c2169c9f1e43d1da390ef4ad2e\"'}"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FILETYPE_FEATURES_FILENAME = os.path.join(ORG + REPO + \"FILETYPE.parquet\")\n",
    "FILETYPE_FEATURES_KEY = os.path.join(s3_input_data_path, ORG, REPO, \"features\")\n",
    "cc.upload_to_ceph(\n",
    "    filetype_keep_cols_df, FILETYPE_FEATURES_KEY, FILETYPE_FEATURES_FILENAME\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "id": "618842b5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.223198Z",
     "iopub.status.busy": "2022-12-07T15:27:48.222752Z",
     "iopub.status.idle": "2022-12-07T15:27:48.231007Z",
     "shell.execute_reply": "2022-12-07T15:27:48.230302Z"
    }
   },
   "outputs": [],
   "source": [
    "# explode the list column so that there is 1 column representing frequency of each word\n",
    "title_wordcount_df = pd.DataFrame(feature_df.title_word_counts_vec.to_list())\n",
    "\n",
    "title_wordcount_df.index = feature_df.index\n",
    "title_wordcount_df.columns = [f\"title_wordcount_{w}\" for w in unique_words]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "id": "457f28ed",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.234488Z",
     "iopub.status.busy": "2022-12-07T15:27:48.234059Z",
     "iopub.status.idle": "2022-12-07T15:27:48.257077Z",
     "shell.execute_reply": "2022-12-07T15:27:48.256166Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title_wordcount_pr               0.412066\n",
       "title_wordcount_workflow         0.298142\n",
       "title_wordcount_functionality    0.298142\n",
       "title_wordcount_add              0.241773\n",
       "title_wordcount_api              0.208804\n",
       "title_wordcount_use              0.206725\n",
       "title_wordcount_inputs           0.206725\n",
       "title_wordcount_srcopsmetrics    0.206725\n",
       "title_wordcount_data             0.206725\n",
       "title_wordcount_pygithub         0.206725\n",
       "title_wordcount_changed          0.206725\n",
       "title_wordcount_testing          0.206725\n",
       "title_wordcount_merge            0.206725\n",
       "title_wordcount_download         0.206725\n",
       "title_wordcount_trigger          0.206725\n",
       "title_wordcount_sample           0.206725\n",
       "title_wordcount_upstream         0.206725\n",
       "title_wordcount_notebook         0.206725\n",
       "title_wordcount_optimization     0.144707\n",
       "title_wordcount_comment          0.144707\n",
       "title_wordcount_inference        0.129352\n",
       "title_wordcount_rate             0.119257\n",
       "title_wordcount_github           0.119257\n",
       "title_wordcount_test             0.110554\n",
       "title_wordcount_text             0.082690\n",
       "title_wordcount_output           0.082690\n",
       "title_wordcount_new              0.082690\n",
       "title_wordcount_yaml             0.076928\n",
       "title_wordcount_handling         0.020672\n",
       "title_wordcount_limit            0.020672\n",
       "title_wordcount_minor            0.020672\n",
       "title_wordcount_updates          0.020672\n",
       "title_wordcount_calls            0.020672\n",
       "title_wordcount_number           0.020672\n",
       "title_wordcount_added            0.020672\n",
       "title_wordcount_training        -0.041345\n",
       "title_wordcount_cleanup         -0.041345\n",
       "title_wordcount_modes           -0.041345\n",
       "title_wordcount_mode            -0.059628\n",
       "title_wordcount_action          -0.087917\n",
       "title_wordcount_basic           -0.103362\n",
       "title_wordcount_update          -0.149071\n",
       "title_wordcount_clean           -0.165380\n",
       "title_wordcount_model           -0.165380\n",
       "title_wordcount_best            -0.165380\n",
       "title_wordcount_feature         -0.165380\n",
       "title_wordcount_notebook3       -0.165380\n",
       "title_wordcount_engineering     -0.165380\n",
       "title_wordcount_choose          -0.165380\n",
       "title_wordcount_fix             -0.193793\n",
       "title_wordcount_sh              -0.219793\n",
       "title_wordcount_entrypoint      -0.219793\n",
       "title_wordcount_yml             -0.227397\n",
       "title_wordcount_train           -0.289414\n",
       "title_wordcount_callingprs      -0.289414\n",
       "title_wordcount_condition       -0.351432\n",
       "title_wordcount_fixed           -0.351432\n",
       "title_wordcount_size            -0.351432\n",
       "title_wordcount_prs             -0.351432\n",
       "dtype: float64"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# compute correlation with ttm\n",
    "corr = title_wordcount_df.corrwith(feature_df[\"ttm_class\"])\n",
    "corr = corr.dropna()\n",
    "corr.sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "id": "0f1495a6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.260974Z",
     "iopub.status.busy": "2022-12-07T15:27:48.260510Z",
     "iopub.status.idle": "2022-12-07T15:27:48.267940Z",
     "shell.execute_reply": "2022-12-07T15:27:48.267183Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "title_wordcount_action          -0.087917\n",
       "title_wordcount_add              0.241773\n",
       "title_wordcount_api              0.208804\n",
       "title_wordcount_basic           -0.103362\n",
       "title_wordcount_best            -0.165380\n",
       "title_wordcount_callingprs      -0.289414\n",
       "title_wordcount_changed          0.206725\n",
       "title_wordcount_choose          -0.165380\n",
       "title_wordcount_clean           -0.165380\n",
       "title_wordcount_cleanup         -0.041345\n",
       "title_wordcount_comment          0.144707\n",
       "title_wordcount_condition       -0.351432\n",
       "title_wordcount_data             0.206725\n",
       "title_wordcount_download         0.206725\n",
       "title_wordcount_engineering     -0.165380\n",
       "title_wordcount_entrypoint      -0.219793\n",
       "title_wordcount_feature         -0.165380\n",
       "title_wordcount_fix             -0.193793\n",
       "title_wordcount_fixed           -0.351432\n",
       "title_wordcount_functionality    0.298142\n",
       "title_wordcount_github           0.119257\n",
       "title_wordcount_inference        0.129352\n",
       "title_wordcount_inputs           0.206725\n",
       "title_wordcount_merge            0.206725\n",
       "title_wordcount_mode            -0.059628\n",
       "title_wordcount_model           -0.165380\n",
       "title_wordcount_modes           -0.041345\n",
       "title_wordcount_new              0.082690\n",
       "title_wordcount_notebook         0.206725\n",
       "title_wordcount_notebook3       -0.165380\n",
       "title_wordcount_optimization     0.144707\n",
       "title_wordcount_output           0.082690\n",
       "title_wordcount_pr               0.412066\n",
       "title_wordcount_prs             -0.351432\n",
       "title_wordcount_pygithub         0.206725\n",
       "title_wordcount_rate             0.119257\n",
       "title_wordcount_sample           0.206725\n",
       "title_wordcount_sh              -0.219793\n",
       "title_wordcount_size            -0.351432\n",
       "title_wordcount_srcopsmetrics    0.206725\n",
       "title_wordcount_test             0.110554\n",
       "title_wordcount_testing          0.206725\n",
       "title_wordcount_text             0.082690\n",
       "title_wordcount_train           -0.289414\n",
       "title_wordcount_training        -0.041345\n",
       "title_wordcount_trigger          0.206725\n",
       "title_wordcount_update          -0.149071\n",
       "title_wordcount_upstream         0.206725\n",
       "title_wordcount_use              0.206725\n",
       "title_wordcount_workflow         0.298142\n",
       "title_wordcount_yaml             0.076928\n",
       "title_wordcount_yml             -0.227397\n",
       "dtype: float64"
      ]
     },
     "execution_count": 75,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr[corr.abs() > 0.04]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "id": "624b3504",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.271629Z",
     "iopub.status.busy": "2022-12-07T15:27:48.270995Z",
     "iopub.status.idle": "2022-12-07T15:27:48.276623Z",
     "shell.execute_reply": "2022-12-07T15:27:48.275850Z"
    }
   },
   "outputs": [],
   "source": [
    "# update list of features to keep for training\n",
    "# take everything that has more than 0.01 correlation magnitude\n",
    "title_word_keep_cols = corr[corr.abs() > 0.04].index.tolist()\n",
    "selected_features += title_word_keep_cols\n",
    "\n",
    "title_wordcount_df = title_wordcount_df[title_word_keep_cols]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "id": "45cd6a78-361c-444f-8065-4127f2e29313",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.280388Z",
     "iopub.status.busy": "2022-12-07T15:27:48.279713Z",
     "iopub.status.idle": "2022-12-07T15:27:48.287740Z",
     "shell.execute_reply": "2022-12-07T15:27:48.287078Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['action',\n",
       " 'add',\n",
       " 'api',\n",
       " 'basic',\n",
       " 'best',\n",
       " 'callingprs',\n",
       " 'changed',\n",
       " 'choose',\n",
       " 'clean',\n",
       " 'cleanup',\n",
       " 'comment',\n",
       " 'condition',\n",
       " 'data',\n",
       " 'download',\n",
       " 'engineering',\n",
       " 'entrypoint',\n",
       " 'feature',\n",
       " 'fix',\n",
       " 'fixed',\n",
       " 'functionality',\n",
       " 'github',\n",
       " 'inference',\n",
       " 'inputs',\n",
       " 'merge',\n",
       " 'mode',\n",
       " 'model',\n",
       " 'modes',\n",
       " 'new',\n",
       " 'notebook',\n",
       " 'notebook3',\n",
       " 'optimization',\n",
       " 'output',\n",
       " 'pr',\n",
       " 'prs',\n",
       " 'pygithub',\n",
       " 'rate',\n",
       " 'sample',\n",
       " 'sh',\n",
       " 'size',\n",
       " 'srcopsmetrics',\n",
       " 'test',\n",
       " 'testing',\n",
       " 'text',\n",
       " 'train',\n",
       " 'training',\n",
       " 'trigger',\n",
       " 'update',\n",
       " 'upstream',\n",
       " 'use',\n",
       " 'workflow',\n",
       " 'yaml',\n",
       " 'yml']"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "title_word_keep_cols = [i[16:] for i in title_word_keep_cols]\n",
    "title_word_keep_cols"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "id": "8b937135-f983-4d6c-961e-f9c7158d108c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.291312Z",
     "iopub.status.busy": "2022-12-07T15:27:48.290908Z",
     "iopub.status.idle": "2022-12-07T15:27:48.296365Z",
     "shell.execute_reply": "2022-12-07T15:27:48.295689Z"
    }
   },
   "outputs": [],
   "source": [
    "title_word_keep_cols_df = pd.DataFrame(title_word_keep_cols, columns=[\"title\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "id": "63f382a4-5826-4d65-aa9f-5f3b56f1aad9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.300161Z",
     "iopub.status.busy": "2022-12-07T15:27:48.299575Z",
     "iopub.status.idle": "2022-12-07T15:27:48.309674Z",
     "shell.execute_reply": "2022-12-07T15:27:48.308943Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>action</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>add</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>api</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>basic</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>best</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>callingprs</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>changed</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>choose</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>clean</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>cleanup</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>comment</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>condition</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>data</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>download</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>engineering</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>entrypoint</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>feature</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>fix</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>fixed</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>functionality</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>github</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>inference</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>inputs</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>merge</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>mode</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>model</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>modes</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>new</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>notebook</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>notebook3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>optimization</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>output</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>pr</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>prs</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>pygithub</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>rate</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>sample</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>sh</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>size</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>srcopsmetrics</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>test</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>testing</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>text</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>train</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>training</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>trigger</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>update</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>upstream</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>use</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>workflow</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>yaml</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>yml</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            title\n",
       "0          action\n",
       "1             add\n",
       "2             api\n",
       "3           basic\n",
       "4            best\n",
       "5      callingprs\n",
       "6         changed\n",
       "7          choose\n",
       "8           clean\n",
       "9         cleanup\n",
       "10        comment\n",
       "11      condition\n",
       "12           data\n",
       "13       download\n",
       "14    engineering\n",
       "15     entrypoint\n",
       "16        feature\n",
       "17            fix\n",
       "18          fixed\n",
       "19  functionality\n",
       "20         github\n",
       "21      inference\n",
       "22         inputs\n",
       "23          merge\n",
       "24           mode\n",
       "25          model\n",
       "26          modes\n",
       "27            new\n",
       "28       notebook\n",
       "29      notebook3\n",
       "30   optimization\n",
       "31         output\n",
       "32             pr\n",
       "33            prs\n",
       "34       pygithub\n",
       "35           rate\n",
       "36         sample\n",
       "37             sh\n",
       "38           size\n",
       "39  srcopsmetrics\n",
       "40           test\n",
       "41        testing\n",
       "42           text\n",
       "43          train\n",
       "44       training\n",
       "45        trigger\n",
       "46         update\n",
       "47       upstream\n",
       "48            use\n",
       "49       workflow\n",
       "50           yaml\n",
       "51            yml"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "title_word_keep_cols_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "2c67ed79-0555-4f0d-9557-3e6d335baeac",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.313022Z",
     "iopub.status.busy": "2022-12-07T15:27:48.312750Z",
     "iopub.status.idle": "2022-12-07T15:27:48.661575Z",
     "shell.execute_reply": "2022-12-07T15:27:48.660729Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ResponseMetadata': {'RequestId': 'lbdt10ni-70m67c-105s',\n",
       "  'HostId': 'lbdt10ni-70m67c-105s',\n",
       "  'HTTPStatusCode': 200,\n",
       "  'HTTPHeaders': {'x-amz-request-id': 'lbdt10ni-70m67c-105s',\n",
       "   'x-amz-id-2': 'lbdt10ni-70m67c-105s',\n",
       "   'access-control-allow-origin': '*',\n",
       "   'access-control-allow-credentials': 'true',\n",
       "   'access-control-allow-methods': 'GET,POST,PUT,DELETE,OPTIONS',\n",
       "   'access-control-allow-headers': 'Content-Type,Content-MD5,Authorization,X-Amz-User-Agent,X-Amz-Date,ETag,X-Amz-Content-Sha256',\n",
       "   'access-control-expose-headers': 'ETag,X-Amz-Version-Id',\n",
       "   'etag': '\"787c188d178b2dffa928c3f901603b10\"',\n",
       "   'date': 'Wed, 07 Dec 2022 15:27:48 GMT',\n",
       "   'keep-alive': 'timeout=5',\n",
       "   'content-length': '0',\n",
       "   'set-cookie': '1a4aa612fe797ac8466d7ee00e5520d5=b9a4e3efe932781180a9a8945a639f83; path=/; HttpOnly; Secure; SameSite=None'},\n",
       "  'RetryAttempts': 0},\n",
       " 'ETag': '\"787c188d178b2dffa928c3f901603b10\"'}"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "TITLE_FEATURES_FILENAME = os.path.join(ORG + REPO + \"titlewords.parquet\")\n",
    "TITLE_FEATURES_KEY = os.path.join(s3_input_data_path, ORG, REPO, \"features\")\n",
    "cc.upload_to_ceph(title_word_keep_cols_df, TITLE_FEATURES_KEY, TITLE_FEATURES_FILENAME)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "105e4b18",
   "metadata": {},
   "source": [
    "Using various feature importance scores such as chi-squared, mutual information statistic, correlation values, we have selected the most relevant features amongst all of our features. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "4cdafdae",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.665869Z",
     "iopub.status.busy": "2022-12-07T15:27:48.665299Z",
     "iopub.status.idle": "2022-12-07T15:27:48.672529Z",
     "shell.execute_reply": "2022-12-07T15:27:48.671742Z"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['size',\n",
       " 'created_at_day',\n",
       " 'created_at_month',\n",
       " 'created_at_weekday',\n",
       " 'created_at_hour',\n",
       " 'changed_files_number',\n",
       " 'body_size',\n",
       " 'commits_number',\n",
       " 'filetype_.ipynb',\n",
       " 'filetype_.yaml',\n",
       " 'filetype_.py',\n",
       " 'filetype_.sh',\n",
       " 'filetype_.yml',\n",
       " 'filetype_.txt',\n",
       " 'filetype_.md',\n",
       " 'filetype_LICENSE',\n",
       " 'filetype_Dockerfile',\n",
       " 'filetype_Containerfile',\n",
       " 'title_wordcount_action',\n",
       " 'title_wordcount_add',\n",
       " 'title_wordcount_api',\n",
       " 'title_wordcount_basic',\n",
       " 'title_wordcount_best',\n",
       " 'title_wordcount_callingprs',\n",
       " 'title_wordcount_changed',\n",
       " 'title_wordcount_choose',\n",
       " 'title_wordcount_clean',\n",
       " 'title_wordcount_cleanup',\n",
       " 'title_wordcount_comment',\n",
       " 'title_wordcount_condition',\n",
       " 'title_wordcount_data',\n",
       " 'title_wordcount_download',\n",
       " 'title_wordcount_engineering',\n",
       " 'title_wordcount_entrypoint',\n",
       " 'title_wordcount_feature',\n",
       " 'title_wordcount_fix',\n",
       " 'title_wordcount_fixed',\n",
       " 'title_wordcount_functionality',\n",
       " 'title_wordcount_github',\n",
       " 'title_wordcount_inference',\n",
       " 'title_wordcount_inputs',\n",
       " 'title_wordcount_merge',\n",
       " 'title_wordcount_mode',\n",
       " 'title_wordcount_model',\n",
       " 'title_wordcount_modes',\n",
       " 'title_wordcount_new',\n",
       " 'title_wordcount_notebook',\n",
       " 'title_wordcount_notebook3',\n",
       " 'title_wordcount_optimization',\n",
       " 'title_wordcount_output',\n",
       " 'title_wordcount_pr',\n",
       " 'title_wordcount_prs',\n",
       " 'title_wordcount_pygithub',\n",
       " 'title_wordcount_rate',\n",
       " 'title_wordcount_sample',\n",
       " 'title_wordcount_sh',\n",
       " 'title_wordcount_size',\n",
       " 'title_wordcount_srcopsmetrics',\n",
       " 'title_wordcount_test',\n",
       " 'title_wordcount_testing',\n",
       " 'title_wordcount_text',\n",
       " 'title_wordcount_train',\n",
       " 'title_wordcount_training',\n",
       " 'title_wordcount_trigger',\n",
       " 'title_wordcount_update',\n",
       " 'title_wordcount_upstream',\n",
       " 'title_wordcount_use',\n",
       " 'title_wordcount_workflow',\n",
       " 'title_wordcount_yaml',\n",
       " 'title_wordcount_yml']"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "selected_features"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ddda4924",
   "metadata": {},
   "source": [
    "## Save Feature Engineered Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "id": "894941f7-9815-44af-867a-9e206502cd7a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.676374Z",
     "iopub.status.busy": "2022-12-07T15:27:48.675748Z",
     "iopub.status.idle": "2022-12-07T15:27:48.692665Z",
     "shell.execute_reply": "2022-12-07T15:27:48.691902Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>size</th>\n",
       "      <th>created_at_day</th>\n",
       "      <th>created_at_month</th>\n",
       "      <th>created_at_weekday</th>\n",
       "      <th>created_at_hour</th>\n",
       "      <th>changed_files_number</th>\n",
       "      <th>body_size</th>\n",
       "      <th>commits_number</th>\n",
       "      <th>filetype_.ipynb</th>\n",
       "      <th>filetype_.yaml</th>\n",
       "      <th>...</th>\n",
       "      <th>title_wordcount_training</th>\n",
       "      <th>title_wordcount_trigger</th>\n",
       "      <th>title_wordcount_update</th>\n",
       "      <th>title_wordcount_upstream</th>\n",
       "      <th>title_wordcount_use</th>\n",
       "      <th>title_wordcount_workflow</th>\n",
       "      <th>title_wordcount_yaml</th>\n",
       "      <th>title_wordcount_yml</th>\n",
       "      <th>ttm_class</th>\n",
       "      <th>time_to_merge</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5</td>\n",
       "      <td>6</td>\n",
       "      <td>12</td>\n",
       "      <td>1</td>\n",
       "      <td>21</td>\n",
       "      <td>2</td>\n",
       "      <td>41</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7</td>\n",
       "      <td>63413.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>3394.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>20</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>12</td>\n",
       "      <td>0</td>\n",
       "      <td>17</td>\n",
       "      <td>2</td>\n",
       "      <td>2</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>9</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>12</td>\n",
       "      <td>3</td>\n",
       "      <td>15</td>\n",
       "      <td>1</td>\n",
       "      <td>53</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>390.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 72 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   size  created_at_day  created_at_month  created_at_weekday  \\\n",
       "0     5               6                12                   1   \n",
       "1     1               5                12                   0   \n",
       "2     0               5                12                   0   \n",
       "3     1               5                12                   0   \n",
       "4     3               1                12                   3   \n",
       "\n",
       "   created_at_hour  changed_files_number  body_size  commits_number  \\\n",
       "0               21                     2         41               1   \n",
       "1               20                     2          1               4   \n",
       "2               20                     1          1               1   \n",
       "3               17                     2          2               3   \n",
       "4               15                     1         53               1   \n",
       "\n",
       "   filetype_.ipynb  filetype_.yaml  ...  title_wordcount_training  \\\n",
       "0                2               0  ...                         0   \n",
       "1                0               1  ...                         0   \n",
       "2                0               0  ...                         0   \n",
       "3                0               1  ...                         0   \n",
       "4                1               0  ...                         0   \n",
       "\n",
       "   title_wordcount_trigger  title_wordcount_update  title_wordcount_upstream  \\\n",
       "0                        0                       1                         0   \n",
       "1                        0                       1                         0   \n",
       "2                        0                       0                         0   \n",
       "3                        0                       0                         0   \n",
       "4                        0                       0                         0   \n",
       "\n",
       "   title_wordcount_use  title_wordcount_workflow  title_wordcount_yaml  \\\n",
       "0                    0                         0                     0   \n",
       "1                    0                         0                     1   \n",
       "2                    0                         0                     0   \n",
       "3                    0                         0                     0   \n",
       "4                    0                         0                     0   \n",
       "\n",
       "   title_wordcount_yml  ttm_class  time_to_merge  \n",
       "0                    0          7        63413.0  \n",
       "1                    0          3         3394.0  \n",
       "2                    0          9            NaN  \n",
       "3                    0          9            NaN  \n",
       "4                    0          1          390.0  \n",
       "\n",
       "[5 rows x 72 columns]"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# join all types of features\n",
    "ttm_dataset = pd.concat(\n",
    "    [feature_df, filetype_df, title_wordcount_df], axis=1, ignore_index=False\n",
    ")\n",
    "\n",
    "# keep only the selected features and the prediction label\n",
    "ttm_dataset = ttm_dataset[selected_features + [\"ttm_class\", \"time_to_merge\"]]\n",
    "\n",
    "ttm_dataset.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "id": "4fd9e422-eaf1-4c80-9375-5d71ac985171",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.696304Z",
     "iopub.status.busy": "2022-12-07T15:27:48.696023Z",
     "iopub.status.idle": "2022-12-07T15:27:48.711170Z",
     "shell.execute_reply": "2022-12-07T15:27:48.710473Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 27 entries, 0 to 26\n",
      "Data columns (total 72 columns):\n",
      " #   Column                         Non-Null Count  Dtype  \n",
      "---  ------                         --------------  -----  \n",
      " 0   size                           27 non-null     int64  \n",
      " 1   created_at_day                 27 non-null     int64  \n",
      " 2   created_at_month               27 non-null     int64  \n",
      " 3   created_at_weekday             27 non-null     int64  \n",
      " 4   created_at_hour                27 non-null     int64  \n",
      " 5   changed_files_number           27 non-null     int64  \n",
      " 6   body_size                      27 non-null     int64  \n",
      " 7   commits_number                 27 non-null     int64  \n",
      " 8   filetype_.ipynb                27 non-null     int64  \n",
      " 9   filetype_.yaml                 27 non-null     int64  \n",
      " 10  filetype_.py                   27 non-null     int64  \n",
      " 11  filetype_.sh                   27 non-null     int64  \n",
      " 12  filetype_.yml                  27 non-null     int64  \n",
      " 13  filetype_.txt                  27 non-null     int64  \n",
      " 14  filetype_.md                   27 non-null     int64  \n",
      " 15  filetype_LICENSE               27 non-null     int64  \n",
      " 16  filetype_Dockerfile            27 non-null     int64  \n",
      " 17  filetype_Containerfile         27 non-null     int64  \n",
      " 18  title_wordcount_action         27 non-null     int64  \n",
      " 19  title_wordcount_add            27 non-null     int64  \n",
      " 20  title_wordcount_api            27 non-null     int64  \n",
      " 21  title_wordcount_basic          27 non-null     int64  \n",
      " 22  title_wordcount_best           27 non-null     int64  \n",
      " 23  title_wordcount_callingprs     27 non-null     int64  \n",
      " 24  title_wordcount_changed        27 non-null     int64  \n",
      " 25  title_wordcount_choose         27 non-null     int64  \n",
      " 26  title_wordcount_clean          27 non-null     int64  \n",
      " 27  title_wordcount_cleanup        27 non-null     int64  \n",
      " 28  title_wordcount_comment        27 non-null     int64  \n",
      " 29  title_wordcount_condition      27 non-null     int64  \n",
      " 30  title_wordcount_data           27 non-null     int64  \n",
      " 31  title_wordcount_download       27 non-null     int64  \n",
      " 32  title_wordcount_engineering    27 non-null     int64  \n",
      " 33  title_wordcount_entrypoint     27 non-null     int64  \n",
      " 34  title_wordcount_feature        27 non-null     int64  \n",
      " 35  title_wordcount_fix            27 non-null     int64  \n",
      " 36  title_wordcount_fixed          27 non-null     int64  \n",
      " 37  title_wordcount_functionality  27 non-null     int64  \n",
      " 38  title_wordcount_github         27 non-null     int64  \n",
      " 39  title_wordcount_inference      27 non-null     int64  \n",
      " 40  title_wordcount_inputs         27 non-null     int64  \n",
      " 41  title_wordcount_merge          27 non-null     int64  \n",
      " 42  title_wordcount_mode           27 non-null     int64  \n",
      " 43  title_wordcount_model          27 non-null     int64  \n",
      " 44  title_wordcount_modes          27 non-null     int64  \n",
      " 45  title_wordcount_new            27 non-null     int64  \n",
      " 46  title_wordcount_notebook       27 non-null     int64  \n",
      " 47  title_wordcount_notebook3      27 non-null     int64  \n",
      " 48  title_wordcount_optimization   27 non-null     int64  \n",
      " 49  title_wordcount_output         27 non-null     int64  \n",
      " 50  title_wordcount_pr             27 non-null     int64  \n",
      " 51  title_wordcount_prs            27 non-null     int64  \n",
      " 52  title_wordcount_pygithub       27 non-null     int64  \n",
      " 53  title_wordcount_rate           27 non-null     int64  \n",
      " 54  title_wordcount_sample         27 non-null     int64  \n",
      " 55  title_wordcount_sh             27 non-null     int64  \n",
      " 56  title_wordcount_size           27 non-null     int64  \n",
      " 57  title_wordcount_srcopsmetrics  27 non-null     int64  \n",
      " 58  title_wordcount_test           27 non-null     int64  \n",
      " 59  title_wordcount_testing        27 non-null     int64  \n",
      " 60  title_wordcount_text           27 non-null     int64  \n",
      " 61  title_wordcount_train          27 non-null     int64  \n",
      " 62  title_wordcount_training       27 non-null     int64  \n",
      " 63  title_wordcount_trigger        27 non-null     int64  \n",
      " 64  title_wordcount_update         27 non-null     int64  \n",
      " 65  title_wordcount_upstream       27 non-null     int64  \n",
      " 66  title_wordcount_use            27 non-null     int64  \n",
      " 67  title_wordcount_workflow       27 non-null     int64  \n",
      " 68  title_wordcount_yaml           27 non-null     int64  \n",
      " 69  title_wordcount_yml            27 non-null     int64  \n",
      " 70  ttm_class                      27 non-null     int64  \n",
      " 71  time_to_merge                  20 non-null     float64\n",
      "dtypes: float64(1), int64(71)\n",
      "memory usage: 15.3 KB\n"
     ]
    }
   ],
   "source": [
    "ttm_dataset.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4a949795",
   "metadata": {},
   "source": [
    "## Save results to Ceph or locally\n",
    "\n",
    "Using the helper function to save the data frame in a parquet format on the Ceph bucket if we are running in automation, and locally if not."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "id": "ecd03940-1b27-4472-ae97-469a4ad8a25a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.715066Z",
     "iopub.status.busy": "2022-12-07T15:27:48.714469Z",
     "iopub.status.idle": "2022-12-07T15:27:48.719508Z",
     "shell.execute_reply": "2022-12-07T15:27:48.718675Z"
    }
   },
   "outputs": [],
   "source": [
    "TTM_DATASET_PATH = os.path.join(s3_input_data_path, ORG, REPO)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "adfdfb1b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-12-07T15:27:48.723125Z",
     "iopub.status.busy": "2022-12-07T15:27:48.722701Z",
     "iopub.status.idle": "2022-12-07T15:27:49.264566Z",
     "shell.execute_reply": "2022-12-07T15:27:49.263493Z"
    }
   },
   "outputs": [],
   "source": [
    "## Sanity check to see if the dataset is the same\n",
    "\n",
    "try:\n",
    "    cc.upload_to_ceph(\n",
    "        ttm_dataset,\n",
    "        TTM_DATASET_PATH,\n",
    "        \"ttm_feature_engineered_dataset.parquet\",\n",
    "    )\n",
    "except ValueError as ve:\n",
    "    print(ve)\n",
    "    print(\"Files already uploaded to S3\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "165910c8",
   "metadata": {},
   "source": [
    "# Conclusion\n",
    "\n",
    "This notebook walked through how to access saved raw Pull Requests data from the previous data collection notebook and we engineered some possible features which can be used to predict time to merge of a PR. \n",
    "\n",
    "The features we engineered are as following:\n",
    "\n",
    "- time_to_merge: time taken to merge a PR in seconds, this is feature we will try to predict\n",
    "- changed_files_number: integer indicating the number of files changed, this is an available column\n",
    "- body_size: integer indicating the number of words in the body of the PR, derived from body\n",
    "- size: integer corresponding to a categorical label generated by GitHub which summarizes the size of the changes made to the code base \n",
    "- title_word_counts_vec: vector or list indicating how many times each word appeared in the title\n",
    "- created_at_day: categorical variable indicating the day of month the PR has been created, derived from created_at\n",
    "- created_at_month: categorical variable indicating the month of the year that the PR has been created, derived from created_at\n",
    "- created_at_weekday: categorical variable indicating the day of the week the PR has been created, derived from created_at\n",
    "- created_at_hour: categorical variable indicating the hour of the day the PR has been created, derived from created_at\n",
    "- commits_number: integer indicating the number of commits in the PR\n",
    "- changed_file_type_vec:  vector or list consisting of number of files changed for the top 100 most frequently occurring data types, derived from changed_files\n",
    "\n",
    "\n",
    "# Next Step\n",
    "\n",
    "In the next [notebook](./03_model_training.ipynb), we will train a classifier to predict time to merge of a PR using the features."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}