Tables module#
The tables module gathers functions designed to show
scores computed from Evaluator objects
in a tabular layout or store them in text files.
Median of station scores#
The function median_station_scores
displays a given list of scores for a given list of
Evaluator objects.
Scores from exceedances contingency tables#
The function exceedances_scores
displays scores from the contingency table of a given list of
Evaluator objects.
Example#
In [1]: import evaltools as evt
In [2]: from datetime import date, timedelta
In [3]: stations = evt.utils.read_listing("./sample_data/listing")
In [4]: start_date = date(2017, 6, 1)
In [5]: end_date = date(2017, 6, 4)
In [6]: observations = evt.Observations.from_time_series(
...: generic_file_path="./sample_data/observations/{year}_no2_{station}",
...: correc_unit=1e9,
...: species='no2',
...: start=start_date,
...: end=end_date,
...: stations = stations,
...: forecast_horizon=2)
...:
In [7]: objs = {}
In [8]: daily_objs = {}
In [9]: for model in ['ENS', 'MFM']:
...: generic_file_path = (
...: f"../doc/sample_data/{model}forecast/"
...: "J{forecast_day}/{year}_no2_{station}"
...: )
...: stations_idx = observations.stations.index
...: simulations = evt.Simulations.from_time_series(
...: generic_file_path=generic_file_path,
...: stations_idx=stations_idx,
...: correc_unit=1,
...: species='no2',
...: model=model,
...: start=start_date,
...: end=end_date,
...: forecast_horizon=2,
...: )
...: objs[model] = evt.Evaluator(observations, simulations)
...: daily_objs[model] = objs[model].daily_max()
...:
In [10]: evt.tables.median_station_scores(
....: objs.values(),
....: forecast_day=0,
....: score_list=['RMSE', 'MeanBias', 'PearsonR'],
....: )
....:
RMSE MeanBias PearsonR
ENS 9.148952 -3.291864 0.506106
MFM 9.871666 -3.013708 0.336417
In [11]: evt.tables.exceedances_scores(
....: daily_objs.values(),
....: forecast_day=0,
....: thresholds=[20, 30, 40],
....: )
....:
threshold = 20
accuracy bias_score success_ratio hit_rate false_alarm_ratio prob_false_detect threat_score equitable_ts peirce_ss heidke_ss rousseau_ss odds_ratio odds_ratio_ss
ENS 0.596154 0.482759 0.785714 0.379310 0.214286 0.130435 0.343750 0.131955 0.248876 0.233146 0.167366 4.074074 0.605839
MFM 0.653846 0.655172 0.789474 0.517241 0.210526 0.173913 0.454545 0.196567 0.343328 0.328551 0.303571 5.089286 0.671554
threshold = 30
accuracy bias_score success_ratio hit_rate false_alarm_ratio prob_false_detect threat_score equitable_ts peirce_ss heidke_ss rousseau_ss odds_ratio odds_ratio_ss
ENS 0.653846 0.181818 1.000 0.181818 NaN NaN 0.181818 0.113636 0.181818 0.204082 0.076923 NaN NaN
MFM 0.692308 0.363636 0.875 0.318182 0.125 0.033333 0.304348 0.184314 0.284848 0.311258 0.250450 13.533333 0.862385
threshold = 40
accuracy bias_score success_ratio hit_rate false_alarm_ratio prob_false_detect threat_score equitable_ts peirce_ss heidke_ss rousseau_ss odds_ratio odds_ratio_ss
ENS 0.692308 NaN NaN NaN NaN NaN NaN 0.000000 0.000000 0.000000 -0.181818 NaN NaN
MFM 0.750000 0.3125 0.8 0.25 0.2 0.027778 0.235294 0.159204 0.222222 0.274678 0.224326 11.666667 0.842105
Out[11]:
[ accuracy bias_score ... odds_ratio odds_ratio_ss
ENS 0.596154 0.482759 ... 4.074074 0.605839
MFM 0.653846 0.655172 ... 5.089286 0.671554
[2 rows x 13 columns],
accuracy bias_score ... odds_ratio odds_ratio_ss
ENS 0.653846 0.181818 ... NaN NaN
MFM 0.692308 0.363636 ... 13.533333 0.862385
[2 rows x 13 columns],
accuracy bias_score ... odds_ratio odds_ratio_ss
ENS 0.692308 NaN ... NaN NaN
MFM 0.750000 0.3125 ... 11.666667 0.842105
[2 rows x 13 columns]]