Download evaltools' sources¶
v1.0.9 (2023-07-25)
v1.0.8 (2023-02-27)
v1.0.7 (2022-02-24)
v1.0.6 (2021-05-19)
v1.0.5 (2021-03-02)
v1.0.4 (2021-01-14)
v1.0.3 (2020-03-12)
v1.0.2
v1.0.1
v1.0.0
HTML Documentation¶
v1.0.9
v1.0.8
v1.0.7
v1.0.6
v1.0.5
v1.0.4
v1.0.3
v1.0.2
v1.0.1
v1.0.0
Release notes¶
What's new in 1.0.9 (2023-07-25)¶
- New score available: kendalltau (Kendall rank correlation coefficient).
- New score available: NRMSE.
- New arguments in plotting.plot_station_scores: ref.
- New arguments in plotting.plot_significant_differences: score_type.
- New arguments in plotting.plot_score_quartiles: score_type.
- New arguments in plotting.plot_station_score_density: score_type.
- New arguments in plotting.plot_station_score_box: score_type.
- New function: tables.average_scores.
- New plotting function: plotting.plot_average_ft_scores.
- New plotting function: fairmode.plot_forecast_target_diagram.
What's new in 1.0.8 (2023-02-27)¶
- New score available: MAE (Mean Absolute Error).
- New plotting function: plotting.plot_data_box.
- New plotting function: plotting.plot_station_score_box.
- New arguments in plotting.plot_values_scatter_plot: xmin, xmax, ymin,
ymax. - New method evaluator.Evaluator.colocate_nan.
- Update FAIRMODE coefficients according to "FAIRMODE Guidance Document on
Modelling Quality Objectives and Benchmarking" version 3.3. - Respecting flake8 conventions in plotting module (this implied the
renaming of some functions, old names are now deprecated). - Objects with hourly time steps are allowed in tables.exceedancesScores.
What's new in 1.0.7 (2022-02-24)¶
- New available scores: success_rate and std(obs-sim).
- New argument in utils.read_listing: decimal.
- New arguments in plotting.plot_timeScores: hourly_timeseries and
start_end. - New module sqlite.
- Fairmode target plot: new argument indicative_color to color legend labels
according to MQI90 and Y90 values. - Fairmode target plot: fix Y90 calculation.
- Fairmode target plot: fix MQI90 calculation when only one station is
taken into account. - All plotting functions now return their matplotlib figures and axes.
What's new in 1.0.6 (2021-05-19)¶
- plotting.plot_dataDensity is now computed only with data
where no missing value is found across all objects. - New argument in plotting.plot_stationScores: marker.
- New module netcdf.
- New method evaluator.Evaluator.remove_negvalues.
- New method evaluator.Evaluator.replace_value.
- Add possibility to use French translation for plotting annotations.
- New plotting function plotting.summary_bar_chart.
- New plotting function plotting.plot_barScoresConc.
- New argument in plotting.plot_stationScores: boundaries_above.
- New argument in plotting.plot_barScores: ref_line.
What's new in 1.0.5 (2021-03-02)¶
- Add possibility to use French translation for plotting annotations.
- New plotting function plotting.summary_bar_chart.
- New plotting function plotting.plot_barScoresConc.
- New argument in plotting.plot_stationScores: boundaries_above.
- New argument in plotting.plot_barScores: ref_line.
What's new in 1.0.4 (2021-01-14)¶
- Add possibility to use data with time step of 3, 6, 8 or 12 hours.
- plotting.plot_medianStationScores now works for objects with daily
time step. - *plotting.plot_timeSeries: median or mean is now performed only with data
where no missing value is found across all objects. - New plotting function plotting.plot_exceedances_scores.
- New arguments in plotting.plot_lineExceedances: black_axes,
nb_of_minor_ticks - New argument in evaluator.Observations.from_cn: correc_unit.
- New argument in evaluator.Simulations.from_cn: correc_unit.
- New arguments in plotting.plot_barExceedances: outputCSV.
- New arguments in plotting.plot_barScores: outputCSV.
- New arguments in plotting.plot_barContingencyTable: outputCSV.
- New arguments in plotting.roc_curve: start_end.
- New arguments in plotting.performance_diagram: start_end.
- New arguments in plotting.plot_timeSeries: min_nb_sta.
- New arguments in plotting.plot_diurnalCycle: nb_of_minor_ticks.
- Various minor fixes.
What's new in 1.0.3 (2020-03-12)¶
- New plotting function plotting.plot_valuesScatterPlot.
- New plotting function plotting.performance_diagram.
- New plotting function plotting.roc_curve.
- New method evaluator.Observations.from_nc.
- New method evaluator.Simulations.from_nc.
- bug fix: legend was sometimes wrong in plot_barScores.
What's new in 1.0.2¶
- New method Evaluator.selectCountries.
- New method Simulations.to_obs().
- New plotting function plotting.plot_barScores.
- New plotting function plotting.plot_barExceedances.
- New plotting function plotting.plot_lineExceedances.
- New plotting function plotting.plot_barContingencyTable.
- New arguments in plotting.plot_stationScores: cmaplabel, extend, land_color, sea_color, mark_by.
- New arguments in plotting.plot_taylorDiagram: norm, frame, crmse_levels.
- New arguments in plotting.plot_timeSeries: thresh, thresh_kw.
- bug fix: a wrong threshold was used in Evaluator.FDscores when multiple days were asked.
- bug fix: variance was used in Taylor diagram instead of standard deviation.
What's new in 1.0.1¶
- evaluator.Observations.movingAverageDailyMax, evaluator.Observations.movingAverageDailyMax and evaluator.Observations.movingAverageDailyMax have no longer keep_first_day argument.
- plotting.plot_stationScores has two new arguments: sea_mask and land_mask.
- New variable plotting.figsize. Its default value is (8, 5) and it corresponds to the value of the argument figsize of matplotlib.pyplot.figure used in evaltools plotting functions. Thus, changing its value in your script allows you to modify the size of all plots you get with evaltools.
- New argument color_by in plotting.plot_comparisonScatterPlot.
- New method evaluator.Evaluator.fairmodeBenchmark.
- New argument obs_style in plotting.plot_timeSeries, plotting.plot_dataDensity and plotting.plot_diurnalCycle.
- New argument nb_of_minor_ticks in plotting.plot_meanTimeScores, plotting.plot_medianStationScores, plotting.plot_timeScores and plotting.plot_timeSeries.
- New methods Simulations.drop_unrepresentative_stations and Observations.drop_unrepresentative_stations.