Skip to content

Commit

Permalink
refactor: Split module 'main' into separate modules (#10)
Browse files Browse the repository at this point in the history 
Closes #8
  • Loading branch information
@a-vester
a-vester authored Sep 22, 2022
1 parent d6095f5 commit 7c11c1f
Show file tree
Hide file tree
Showing 10 changed files with 1,611 additions and 1,460 deletions.
63 changes: 39 additions & 24 deletions src/pyratings/__init__.py
View file
Original file line number Diff line number Diff line change
@@ -1,31 +1,46 @@
"""
Copyright 2022 HSBC Global Asset Management (Deutschland) GmbH
# Copyright 2022 HSBC Global Asset Management (Deutschland) GmbH
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
"""
from pyratings.main import (
get_pure_ratings,
from pyratings.aggregate import get_weighted_average
from pyratings.clean import get_pure_ratings
from pyratings.consolidate import (
get_best_ratings,
get_second_best_ratings,
get_worst_ratings,
get_scores_from_ratings,
get_scores_from_warf,
get_ratings_from_scores,
get_ratings_from_warf,
get_warf_from_scores,
get_warf_from_ratings,
get_weighted_average,
get_warf_buffer,
)
from pyratings.get_ratings import get_ratings_from_scores, get_ratings_from_warf
from pyratings.get_scores import get_scores_from_ratings, get_scores_from_warf
from pyratings.get_warf import get_warf_from_ratings, get_warf_from_scores
from pyratings.utils import (
_assert_rating_provider,
_extract_rating_provider,
_get_translation_dict,
)
from pyratings.warf import get_warf_buffer

# define public functions
__all__ = [
"get_best_ratings",
"get_pure_ratings",
"get_ratings_from_scores",
"get_ratings_from_warf",
"get_scores_from_ratings",
"get_scores_from_warf",
"get_second_best_ratings",
"get_warf_buffer",
"get_warf_from_ratings",
"get_warf_from_scores",
"get_weighted_average",
"get_worst_ratings",
]
66 changes: 66 additions & 0 deletions src/pyratings/aggregate.py
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,66 @@
# Copyright 2022 HSBC Global Asset Management (Deutschland) GmbH
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import pandas as pd


def get_weighted_average(data: pd.Series, weights: pd.Series) -> float:
"""
Computes weighted average.
Parameters
----------
data
Contains numerical values.
weights
Contains weights (between 0 and 1) with respect to data.
Returns
-------
float
Weighted average data.
Notes
-----
Computing the weighted average is simply the sumproduct of `data` and `weights`.
``nan`` in `data` will be excluded from calculating the weighted average. All
corresponding weights will be ignored. As a matter of fact, the remaining
weights will be upscaled so that the weights of all ``non-nan`` rows in `data` will
sum up to 1 (100%).
Examples
--------
>>> import numpy as np
>>> import pandas as pd
>>> rtg_scores = pd.Series(data=[5, 7, 9])
>>> wgt = pd.Series(data=[0.5, 0.3, 0.2])
>>> get_weighted_average(data=rtg_scores, weights=wgt)
6.4
>>> warf = pd.Series(data=[500, 735, np.nan, 93, np.nan])
>>> wgt = pd.Series(data=[0.4, 0.1, 0.1, 0.2, 0.2])
>>> get_weighted_average(data=warf, weights=wgt)
417.29
"""
# find indices in warf that correspond to np.nan
idx_nan = data[pd.isna(data)].index

# sum weights of securities with an actual rating, i.e. rating is not NaN
weights_non_nan = 1 - sum(weights.loc[idx_nan])

# upscale to 100%
weights_upscaled = weights / weights_non_nan

return data.fillna(0).dot(weights_upscaled)
130 changes: 130 additions & 0 deletions src/pyratings/clean.py
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,130 @@
# Copyright 2022 HSBC Global Asset Management (Deutschland) GmbH
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# https://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from typing import Union

import pandas as pd


def get_pure_ratings(
ratings: Union[str, pd.Series, pd.DataFrame]
) -> Union[str, pd.Series, pd.DataFrame]:
"""Removes rating watches/outlooks.
Parameters
----------
ratings
Rating may contain watch, such as `AA- *+`, `BBB+ (CwNegative)`.
Outlook/watch should be seperated by a blank from the actual rating.
Returns
-------
Union[str, pd.Series, pd.DataFrame]
String, Series, or DataFrame with regular ratings stripped off of watches.
The name of the resulting Series or the columns of the returning DataFrame will
be suffixed with `_clean`.
Examples
--------
Cleaning a single rating:
>>> get_pure_ratings("AA- *+")
'AA-'
>>> get_pure_ratings("Au")
'A'
Cleaning a `pd.Series`:
>>> import numpy as np
>>> import pandas as pd
>>> rating_series=pd.Series(
... data=[
... "BB+ *-",
... "BBB *+",
... np.nan,
... "AA- (Developing)",
... np.nan,
... "CCC+ (CwPositive)",
... "BB+u",
... ],
... name="rtg_SP",
... )
>>> get_pure_ratings(rating_series)
0 BB+
1 BBB
2 NaN
3 AA-
4 NaN
5 CCC+
6 BB+
Name: rtg_SP_clean, dtype: object
Cleaning a `pd.DataFrame`:
>>> rtg_df = pd.DataFrame(
... data={
... "rtg_SP": [
... "BB+ *-",
... "BBB *+",
... np.nan,
... "AA- (Developing)",
... np.nan,
... "CCC+ (CwPositive)",
... "BB+u",
... ],
... "rtg_Fitch": [
... "BB+ *-",
... "BBB *+",
... pd.NA,
... "AA- (Developing)",
... np.nan,
... "CCC+ (CwPositive)",
... "BB+u",
... ],
... },
... )
>>> get_pure_ratings(rtg_df)
rtg_SP_clean rtg_Fitch_clean
0 BB+ BB+
1 BBB BBB
2 NaN <NA>
3 AA- AA-
4 NaN NaN
5 CCC+ CCC+
6 BB+ BB+
"""
if isinstance(ratings, str):
ratings = ratings.split()[0]
ratings = ratings.rstrip("uU")
return ratings

elif isinstance(ratings, pd.Series):
# identify string occurrences
isstring = ratings.apply(type).eq(str)

# strip string after occurrence of very first blank and strip character 'u',
# which has usually been added without a blank
ratings[isstring] = ratings[isstring].str.split().str[0]
ratings[isstring] = ratings[isstring].str.rstrip("uU")
ratings.name = f"{ratings.name}_clean"
return ratings

elif isinstance(ratings, pd.DataFrame):
# Recursive call of `get_pure_ratings`
return pd.concat(
[get_pure_ratings(ratings=ratings[col]) for col in ratings.columns], axis=1
)
Loading

0 comments on commit 7c11c1f

Please sign in to comment.