Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Update develop-ref after #1892 #1897

Merged
merged 144 commits into from
Aug 31, 2021
Merged

Update develop-ref after #1892 #1897

merged 144 commits into from
Aug 31, 2021

Conversation

JohnHalleyGotway
Copy link
Collaborator

@JohnHalleyGotway JohnHalleyGotway commented Aug 31, 2021

Pull Request Testing

Pull Request #1892 for Issue #1788 modifies the existing GO Index output and adds 2 new STAT-Analysis output files for the SSIDX line type. The modified and new files are flagged as a difference in the nightly build output. Updating develop-ref to get past this change.

  • Describe testing already performed for these changes:

    Inspected 20210830 NB output on kiowa. Here are the 2 new output files:
ERROR: folder MET-develop-ref/test_output missing 2 files
    ref_config/stat_analysis/go_index_by_option.stat 
    ref_config/stat_analysis/sfc_ss_index_by_option.stat 

And the modified output file still computes the same GO Index value. Its just formatted differently:

COMPARING ref_config/stat_analysis/go_index.out
< COL_NAME: GO_INDEX
< GO_INDEX:  0.99873
> COL_NAME:        FCST_MODEL           REF_MODEL N_INIT N_TERM N_VLD SS_INDEX
>  GO_INDEX AFWAv3.4_Noahv3.3 AFWAv3.4_Noahv2.7.1      1     48    48  0.99873

Confirmed that these are the only diffs.

  • Recommend testing for the reviewer(s) to perform, including the location of input datasets, and any additional instructions:

    None

  • Do these changes include sufficient documentation updates, ensuring that no errors or warnings exist in the build of the documentation? [Yes]

  • Do these changes include sufficient testing updates? [Yes]

  • Will this PR result in changes to the test suite? [Yes]

    If yes, describe the new output and/or changes to the existing output:

    Update reference version, as expected.

  • Please complete this pull request review by [Fill in date].

Pull Request Checklist

See the METplus Workflow for details.

  • Complete the PR definition above.
  • Ensure the PR title matches the feature or bugfix branch name.
  • Define the PR metadata, as permissions allow.
    Select: Reviewer(s)
    Select: Organization level software support Project or Repository level development cycle Project
    Select: Milestone as the version that will include these changes
  • After submitting the PR, select Linked issues with the original issue number.
  • After the PR is approved, merge your changes. If permissions do not allow this, request that the reviewer do the merge.
  • Close the linked issue and delete your feature or bugfix branch from GitHub.

jprestop and others added 30 commits June 14, 2021 12:24
Commented out items and added text to description
…to_laton

Bugfix 1817 point2grid latlon to laton
…arsing

Bugfix 1508 tc_gen file list parsing
…_double_type_lat_lon

Bugfix 1838 point2grid supports the double type lat/lon
* Per #1833, changed references to met_help to Discussions.

* Per #1833, fixed typo and removed an unnecessary word
Howard Soh and others added 26 commits July 30, 2021 15:51
* Per #1864, remove stale GitHub pages URL for the MET User's Guide.

* Per #1864, found stale GitHub IO links in the issue templates. Should make this same fix in the develop and default branches for all the METplus repos.

* Per #1864, fix stale GitHub IO link in the PR template.
* testing possible new drop-down menus for FAQ

* adding sphinx-panels to extensions to be able to use drop down menus in appendixA

* Per #1834, adding sphinx-panels to requirements.txt to see if that fixes the error

* Per #1834, added sphinx.ext to beginning of sphinx_panels to see if that fixes the error

* Per #1834, remove sphinx.ext and add trailing comma to end of extensions to see if that fixed the error

* Per #1834, added sphinx-panels to pip install command to see if that fixes the error

* Per #1834, added extra space in comment to get docs to build to see if change to documentation.yml will fix an error

* Per #1834, added return character to get rid of warning

* testing dropdown menus #1

* testing dropdown menus #2 removing lines and dotted lines

* testing #3

* testing 4

* fixing numbering, removing 2nd pulldown menu, trying to fix box width

* adding code block in dropdown

* trying to fix width

* trying to fix width with spacing

* trying to fix width with spacing 2nd dropdown

* removing panels

* solved problem, just use dropdown no panels

* adding second dropdown menu

* removing the dropdown menus and fixing the spacing.

* fixing typos

* adding in matched pairs

* adding in formating files for NetCDF and making FILE_IO a header

* fixing spacing

* time slice

* fixing spacing

* UNIX time conversion

* fixing spacing

* fixed-width

* scientific notation

* removing which per Julie P

* adding Gen_Vx_Mask section

* fixing formatting

* changing to numbers

* complex masking region

* neighborhood methods boundaries

* Neighborhood Methods to Compute Fraction

* adding Grid_Stat - How do I use Config File Setup to Read a NetCDF file

* using backticks with asterics

* one more backtick with asterics

* Grid_Stat - What would be an example of Verifying Probabilities

* fixing typos

* What is an example of using Grid-Stat with Regridding and Masking Turned On

* fixing typos

* Grid_Stat - What would be an example of Verifying Probabilities Example2

* trying to fix typos

* trying to fix code-block spacing

* Per #1834, attempting to fix warning.

* How do I use different masks in MET tools using MODE as an example

* trying to fix bullet list

* trying to fix bullet list 2

* trying to fix bullet list 2

* trying to fix bullet list 3

* Per #1834, attempting formatting changes

* Per #1834, attempting formatting changes

* Per #1834, attempting formatting changes

* Per #1834, attempting formatting changes

* Per #1834, attempting formatting changes

* Per #1834, modified formatting, including adding backslashes and removing extra returns in command line calls, and adding indentation for config file notation, and adding a return after each 'A.'

* Per #1834, changed code-block from ini to none to fix formatting

* *Pcp-Combine - What are some examples using -add

* Pcp-Combine -  How do I add and subtract with Pcp-Combine

* Pcp-Combine - How do I Combine 12-hour Accumulated Precipitation from Two Different Initialization Times

* Pcp-Combine - How Do I Correct a Precipitation Time Range

* Pcp-Combine - What Data Formats does MET Read

* Pcp-Combine - How does pass through work

* Pcp_Combine - How do I use -pcprx to run a project faster

* Pcp-Combine - How Do I Enter the Time Format Correctly

* Pcp-Combine - How do I use -subtract

* fixing formatting

* Pcp_Combine - How Do I Use -sum, -add, and -subtract to Achieve Same Accumulation Interval

* Pcp-Combine - What is the difference between -sum vs. -add

* Pcp-Combine - How Do I Select a Specific GRIB Record to Use

* fixing spacing for original text

* Plot-Data-Plane - How Do I Inspect Gen_Vx_Mask Output

* fixing typo

* Plot-Data-Plane - How Do I Specify GRIB Version

* Plot-Data-Plane - last 2 entries

* 2 misc. entries

* fixing typo

* Stat_Analysis - How does -aggregate_stat work

* Stat_Analysis - remaining entries

* fixing typo

* tc-pairs

* tc-stat

* troubleshooting area

* another stat-analysis

* another stat-analysis

* utilities

* fixing dashes

* Per #1834, trying out syntax

* Per #1834, removing added syntax

* fixing returns with bolding

* Per #1834, modified formatting

* adding italics for directories #1834

* replacing code-block ini with none to remove red lettering

* fixing questions to lower case #1834

* linking to section using numref #1834

* linking to section using numref #1834 attempt 2

* linking to section using numref #1834 attempt 3

* linking to section using numref #1834 attempt 4

* linking to section using numref #1834 attempt 5

* linking to section using numref #1834 attempt 6

* linking to section using numref #1834 attempt 7

* linking to section using numref #1834 attempt 8

* linking to anothre section #1834

* linking to another section #1834

* removing old info #1834

* grammar updates #1834

* grammar updates #1834

* fixing typos #1834

* email changes with Julie #1834

* trying to fix warning #1834

* changes #1834

* Per #1834, updating question titles and answers

* Per #1834, made updates to the content.

* Per #1834, made updates to the content

* Per #1834, made updates to the content

* Per #1834, made some major changes to the first 3 FAQ's. I don't have the time/bandwidth to review these completely at this time, but wanted to include at least these updates.

* testing some of the updates John O. suggested.

* updates John O. suggested take 2 #1834.

* updates John O. suggested take 3 #1834.

* updates John O. suggested take 4 #1834.

* updates John O. suggested take 5 #1834.

* updates John O. suggested.  Some didn't show up even though they've been removed #1834.

* trying to fix numbered list indenting #1834

* fixing numbered list indenting #1834

* cleaning up formatting #1834

* fixing broken links #1834

Co-authored-by: Julie.Prestopnik <[email protected]>
Co-authored-by: John Halley Gotway <[email protected]>
* Per #1875, fix set_cdf_ta() to initialize cdf_ta prior to populating it.

* Per #1875, add debug(4) log messages whenever defining a new aggregation case in stat-analysis.

* Per #1875, switch log messages about new cases from Debug(4) to Debug(3).
* Per #1673, add a distance_map.beta_value entry to all the Grid-Stat config files.

* Per #1673, enhance Grid-Stat to parse the new beta_value option and set its value in the library code.

* Per #1673, set the user-defined beta_value when processing the distance map scores.

* Per #1673, add 4 new DMAP columns write them. Just writing bad data values right now, still need to actually compute them! Also need to update Stat-Analysis to update the parsing/aggregation of these new values.

* Per #1673, enable beta_value to be set to a default bad data value.

* Per #1673 and @ericgilleland instruction (#1673 (comment)), strip asymmetric G-Beta (AGBETA) from the output.

* Per #1673, fix typo that was causing the code not to compile.

* Per #1673, update MET User's Guide about G/GBETA.

* Per #1673, report BETA_VALUE = TOTAL / 2 in the output instead of NA, as directed by Eric.

* Per #1673, remove descriptions of Grid-Stat-specific config options for nbrhd, fourier, gradient, and distance_map from the overview section since they belong in the Grid-Stat chapter. Also add direct links for them in the Grid-Stat chapter.

* Per #1673, do not reset the beta_value in DMAP::clear().

* Per #1673, log the DMAP options only after the beta_value has been determined.

* Per #1673, correct the default beta_value setting as N*N / 2.0.

* Per #1673, trying to format G-beta well.

* Per #1673, trying to format G-beta well.

* Per #1673, beta is formatted as \beta and not \Beta in latex.

* Per #1673, beta is formatted as \beta and not \Beta in latex.

* Per #1673, switch to defining beta_value as a function of the input grid size.

* Per #1673, push initial appendix C updates for GBETA.

* Per #1673, working on G and GBETA equations.

* Per #1673, working on G and GBETA equations.

* Per 1673, update GBETA details in appendix C.

* Per #1673, add Eric's info about selecting beta.

* Per #1673, update reference for Gilleland-2019 to Gilleland_PartI-2020 and Gilleland_PartII-2020.

* Per 1673, fix typo.

* Per #1673, making changes requested by Eric.

* Per #1673, making changes requested by Eric.

* Per #1673, update references based on Eric's feedback.

* Update appendixD.rst (#1869)

Updated this appendix.

* Update appendixD.rst

* Per #1673, making the formatting of Gilleland-2020 part I and II references consistent.

* Per #1673, update Grid-Stat docs to clarify that GBETA is only computed on the FULL verification domain and not any masking regions.

* Per #1673, add DataPlane::n_good_data() function to return a count valid data values.

* Per #1673, update vx_statistics library code to only set gbeta/beta_value to bad data for sub-domains.

* Per #1673, update grid_stat to mask out bad data between the fcst/obs fields before computing distance maps. This will change existing unit test results! Also store the count of valid data values in the full verification domain.

* Per #1673, since GBETA is only reported when VX_MASK = FULL, updating GridStatConfig_all to include the FULL masking region to make this example more meaningful.

Co-authored-by: ericgilleland <[email protected]>
* Per #1764 added write_orank_row() to do_hera_ens(). SL

* Per issue #1764, added orank stat type for ensemble output. SL

* Per issue #1764, added code to setup_txt_files() for ORANK output. SL

* Added some comments for debuging. SL

* Per #1764, add orank to the list of outputs created by point_stat.

* Per #1764, add functions to compute the number of requested HiRA probabilities and ensemble members. Also update n_txt_row() to keep track of the number of ORANK rows to be written. Not totally sure if I'm counting these ORANK output lines correctly though.

* Per #1764, update logic in setup_txt_files() to call get_max_n_hira_prob() and get_max_n_hira_ens() functions.

* Per #1764, no real changes. Just fixing spacing.

* Per issue #1764, cleaned up print (cout) debug statements. SL

* Per issue #1764 added 'orank = NONE' to the config files for compatibility with changes to point_stat for calculating ORANK. SL

* Per issue #1764 added 'orank = STAT' to config file to produce ORANK output lines in point_stat. SL

* Per issue #1764 added content in all relevant sections for ORANK: observation rank statistic. SL

* Added 'orank = NONE' to config file for compatibility with changes to point_stant for ORANK. SL

* Added 'orank = NONE' to config file for compatibility with changes to point_stant for ORANK. SL

* Apply suggestions from code review

Co-authored-by: George McCabe <[email protected]>

* Per #1764, no real changes, just fixing indents.

* Per #1764, no real changes, just fixing indents.

* Per #1764, no real changes, just fixing indents.

* Per #1764, putting the docs for VCNT back into its logical order with the other vector line types. The order that they're listed here does not need to match their order in the enumeration in the code.

* Per #1764, fixing a vcnt typo I found in the grid-stat chapter... and making them consistent.

* Per issue #1764: After the call to write_orank_row, added lines to reset the observation begin and end times. SL

* Per #1764, no code changes, just fixing indents.

Co-authored-by: Seth Linden <[email protected]>
Co-authored-by: John Halley Gotway <[email protected]>
Co-authored-by: John Halley Gotway <[email protected]>
Co-authored-by: George McCabe <[email protected]>
* Per #1886, port over the bugfixes from main_v10.0 to develop.

* Per #1886, make grid_diag error out if to_grid is set to FCST or OBS.
* Per #1870. add TCPairs config entries for valid_inc, valid_exc, and write_valid.

* Per #1870, removing 2 files I accidentally added.

* Per #1870, add code to parse the new tc_pairs config options.

* Per #1870, unrelated... just fixing spacing.

* Per #1870, update TrackPairInfo/Array classes and tc_pairs application code to use the write_valid option to subset the tracks by requested valid time before writing them.

* Per #1870, update the User's Guide with the new config options. Document valid_inc/valid_exc in the common TC config section and add write_valid to the TC-Pairs chapter.

* Per #1870, move the subset_write_valid() functionality to the library code and support it for both track and probability arrays.

* Per #1870, update unit_tc_pairs.xml with 2 new calls to exercise the new write_valid functionality for both tracks and probabilities.

* Per #1870, correct some typos in the docs... PROBRI and probri should be PROBRIRW and probrirw, resp.

* Per #1870, switching match_points from TRUE back to FALSE after I acciedentally committed that change.

* Per #1870, just fixing indent.

Co-authored-by: John Halley Gotway <[email protected]>
* Per #1788, add initial definition of the SSIDX line type.

* Per #1788, add a STAT-Analysis config file for computing the CBS Score. This is currently just a copy of the GO Index.

* Per #1788, add hooks for computing -job cbs_score.

* Per #1788, since we're adding a new SSIDX line type, the columns for that line type replace the job_ss_columns and job_go_columns arrays.

* Per #1788, rename cbs_score to cbs_index per UK Met Office instruction on 7/22/21. This convention is also consistent with the existing go_index and ss_index jobs.

* Per #1788 rename STATAnalysisConfig_CBS_Score to STATAnalysisConfig_CBS_Index

* Per #1788, update Makefile to handle name change.

* Per #1788, add support for SSIDX stat line type.

* Per #1788, define 4 output SSIDX columns.

* Per #1788, define an SSIndexInfo struct for stashing/writing SS Index values.

* Per #1788, write the SSIDX columns. However we don't need to function to write the whole row since it's only written by Stat-Analysis.

* Per #1788, create an AggrSSIndexInfo struct to keep track of unique stat header column input values.

* Per #1788, update the stat-analysis job code to make use of the new structs.

* Per #1788 fix typo.

* Per #1788, remove GO Index defintion from the code comments. That detail belongs elsewhere.

* Per #1788, since coding up the GO Index several CNT and CTC columns have been added. Adding them here to let them be included as part of the Skill Score index computations.

* Per #1788, update check_hdr_str() to print the name of the problematic header columns.

* Per #1788, add a set(one) member function to multiple array classes. This calls clear or erase on the array prior to adding a single element. Add this for StringArray, NumArray, and ThreshArray, and CRCArray.

* Per #1788, update the compute_ss_index() function in stat_analysis to make the code more concise by calling the newly adding singular set functions for arrays.

* Per #1788, add FCST_MODEL and REF_MODEL output columns to the SSIDX line type.

* Per #1788, add a new N_INIT column to the SSIDX line type to report the number of initialization times represented by this skill score. Also define the SSIDXData struct to store the skill score index job output.

* Per #1788, update STATAnalysisJob to explicitly set the job type directly rather than parsing a string.

* Per #1788, major change here adding skill_score_index_job.h/.cc to handle the computation of the skill score indices.

* Per #1788, update stat_analysis source code to make use of the new classes in skill_score_index_job.h/.cc.

* Per #1788, write a log message listing the model initialization times over which the skill score index is computed.

* Per #1788, tweak log message.

* Per #1788, compute n_term as the max lenght of the required arrays.

* Per #1788, drop in the ACTUAL definition of the GO index provide by the DTC AF project team. Also simplify the existing arrays down to lenght one if the value remains constant.

* Per #1788, correct CBS Index config file. Some parameters had 40 elements while others only had 32.

* Per #1788, do not write SSIDX output if there are no valid terms found.

* Per #1788, format the values in the skill score index log messages, replacing -9999 with NA.

* Per #1788, update the get_stat() member functions. For 0 pairs, return bad data for all statistics other than TOTAL.

* Per #1788, working on readability of the warning and error messages.

* Per #1788, define the ss_index_name directly in the STATAnalysis config file to make it really easy to modify.

* Per #1788, store the ss_index_name configuration option in the StatJob class. Also support a command line -ss_index_name option to override the config file value.

* Per #1788, change from name to ss_index_name variable to be a bit more descriptive in the variable names.

* Per #1788, since ss_index_name is in the StatJob class we no longer need to pass it around as an argument.

* Per #1788, simply the skill score index job handling logic a bit. We only need to update the default job once, not twice.:

* Revert "Per #1788, since ss_index_name is in the StatJob class we no longer need to pass it around as an argument."

This reverts commit cde0be0.

* Revert "Revert "Per #1788, since ss_index_name is in the StatJob class we no longer need to pass it around as an argument.""

This reverts commit bbfd838.

* Per #1788, I'd pruned too much code from stat_analysis.cc which caused the processing of config file jobs to fail. Fixing that here.

* Per #1788, switch Z0 to L0 to match the sample data for computing CBS Index.

* Per #1788, make it so that cbs_index and go_index work from config file.

* Per #1788, update the documentation for the skill score index changes.

* Per #1788, just move some variable definitions further down since they're only used sometimes.

* Per #1788, make job output description into a numbered section.

* Per #1788, little tweak.

* Per #1788, update the STATAnalysisConfig files to add ss_index_name and hss_ec_value, as needed.

* Per #1788, add new unit tests to exercise recent ss_index changes and update the Stat-Analysis documentation.

* Update met/docs/Users_Guide/stat-analysis.rst

Co-authored-by: jprestop <[email protected]>

* Update met/docs/Users_Guide/stat-analysis.rst

Sorry for the typos! Thanks for catching them.

Co-authored-by: jprestop <[email protected]>

* Update met/docs/Users_Guide/stat-analysis.rst

Co-authored-by: jprestop <[email protected]>

* Per #1788, fix typo in error message.

* Per #1788, fix a bug for ss_index... changing 'add' to 'set'. Also set the line_type seperately for each term, if requested.

* Per #1788, tweak the ss_index logic slightly to make it more efficient. For each input STATLine, instead of checking all the index terms, stop searching after the first match. This assumes that each input line should be used for only one of the terms... and I think that's a pretty reasonable assumption.

* Per #1788, minor tweaks to the STAT-Analysis docs.

* Per #1788, add ss_index_vld_thresh option to existing STAT-Analysis config files.

* Per #1788, add StatAnalysisJob ss_index_vld_thresh entry and parse it from the STAT-Analysis config file.

* Per #1788, update stat_analysis code to check ss_index_vld_thresh setting when computing the skill score index. Also update the Stat-Analysis chapter of the User's Guide.

* Per #1788, tweak one test to exercise a non-default ss_index_vld_thresh value.

Co-authored-by: jprestop <[email protected]>
@JohnHalleyGotway JohnHalleyGotway added this to the MET 10.1.0 milestone Aug 31, 2021
@JohnHalleyGotway JohnHalleyGotway merged commit 875f867 into develop-ref Aug 31, 2021
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

7 participants