stats.errors.error_summary

stats.errors.error_summary(errors, *, quantiles=(0.05, 0.95))

Summarise the distribution of forecast errors, one column per entry.

For each entry the summary reports mean, median, the requested quantiles, standard deviation (ddof=1), minimum, and maximum — the row layout of a manuscript error-statistics table.

Parameters

Name	Type	Description	Default
errors	Mapping[str, pd.Series]	Mapping of entry name to its error series (`forecast - actual`).	required
quantiles	tuple[float, …]	Quantile levels to include, each in `(0, 1)`. Defaults to `(0.05, 0.95)`.	`(0.05, 0.95)`

Returns

Name	Type	Description
	pd.DataFrame	pd.DataFrame: Indexed by statistic name
	pd.DataFrame	(`["mean", "median", *[f"q{q:g}"], "std", "min", "max"]`), one
	pd.DataFrame	column per entry, in mapping order.

Raises

Name	Type	Description
	TypeError	When an entry value is not a `pd.Series`.
	ValueError	When `errors` is empty or a quantile is outside `(0, 1)`.

Examples

import pandas as pd
from spotforecast2_safe.stats.errors import error_summary

errors = {"model": pd.Series([-2.0, -1.0, 0.0, 1.0, 2.0])}
table = error_summary(errors)
print(table.round(2).to_string())
assert table.loc["mean", "model"] == 0.0
assert table.loc["min", "model"] == -2.0

           model
statistic       
mean        0.00
median      0.00
q0.05      -1.80
q0.95       1.80
std         1.58
min        -2.00
max         2.00