2. METcalcpy develop Release Notes

When applicable, release notes are followed by the GitHub issue number which describes the bugfix, enhancement, or new feature: METcalcpy GitHub issues.

2.1. Version develop release notes (2021-05-10)

2.1.1. Version 1.1.0 release notes (20210511)

New Functionality:

  • Add OMI and RMM statistics (#89)

  • Add calculation of CTC statistics (#77)

  • Add sorting of CTC dataframe by fcst_thresh (#75)

Enhancements:

Internal:

  • Add bootstrap package classes (#96)

  • Add pingouin package classes (#98)

Bugfixes:

  • Fixed CTC statistics for the ROC diagram (#77)

2.1.2. Version 1.0.0 release notes (20210511)

New Functionality:

  • Add calculations for Space-time Coherence Spectra from NOAA PSL Diagnostics Package (#29)

  • Utilize METdbload .stat file reader utilities to allow METcalcpy to read these files (#31)

  • Utilize METdbload .stat file reader utilities to allow METcalcpy to read these files (#31)

  • Create a template for reading in a data file (of raw data) and create a data file reader (#7)

  • Add new ECNT stats to aggregation logic (#69)

  • Implement a method that returns specified column values (#73)

  • Support for GridDiag analysis (#53)

  • Calculation of CTC stats added: LODDS, ODDS, ORSS, SEDI, SEDS, EDI, EDS stats to the ctc_statistics module** (#60)

  • Compute pairwise differences for “group” statistics (#13)

  • Add equivalence testing interval bounds to the existing continuous line type. Two fields are added:lower bound and upper bound (similar to calculating a normal confidence interval except it requires using the non-central t-distribution (#1)

  • Add calculation for Difficulty Index from NRL (#88)

  • Enhance bootstrapping to support circular temporal block bootstrap with overlapping blocks (#3)

  • Add equivalence testing interval bounds to the existing continuous line type. Two fields are added:

    lower bound and upper bound (similar to calculating a normal confidence interval except it requires using the non-central t-distribution (#1)

  • Add calculation for Difficulty Index from NRL (#30)

  • Create coordinate converter for [0,360] to [-180,180] (#21)

  • Add calculations used for Hovmoller Diagram, as contributed from the NOAA PSL diagnostics package (#28)

  • Create a method for creating series permutations that creates results consistent with R (#44)

  • Add calculation for Difficulty Index from NRL** (#30)

  • Create coordinate converter for [0,360] to [-180,180] (#21)

Enhancements:

  • Documentation added (#6)

  • Replicate METviewer Reliability diagram using Python (#48)

  • Provide support for performing vertical interpolation of fields between grids with pressure and height coordinates (#20)

  • Incorporate the calculation of the Difficulty Index from Naval Research Lab (#27)

Internal:

  • Rename boot Confidence Interval’s (CI) columns (#79)

  • Initial design for data input logic to be used by METplotpy and METviewer (#8)

  • UML diagram of design for data input logic (#64)

  • Developer tests are working (#65)

  • Fix typo in setup.py (#43)

  • Confirm that Event Equalization for off-setting initialization is still working (#16)

  • Move convert_lons_indices() function from plot_blocking.py to utils.py in METcalcpy (#33)

  • Fix a typo (misspelling of package name) in code that creates packaging (#43)

  • Move convert_lons_indices() function from plot_blocking.py to utils.py in METcalcpy (#33)

Bugfixes:

  • Fix event equalization and agg_stat logic (#52)

  • Fix the No-Skill reference line on Reliability Plot implemented in Python (#26)

  • Fix the calculation of ECNT_RMSE statistic so the Python and R implementations are consistent (#42)

  • Change the default setting of calc_difficulty_index to reproduce results originally generated from Naval Research Lab (#37)

  • Fix the No-Skill reference line on Reliability Plot implemented in Python (#26)

  • Fix the calculation of ECNT_RMSE statistic so the Python and R implementations are consistent (#42)

  • Change the default setting of calc_difficulty_index to reproduce results originally generated from Naval Research Lab (#37)

  • Fix event equalization and agg_stat logic (#52)