`digger.compare.confusion`¶

Module Contents¶

Functions¶

`calculate_metrics`	Calculate binary classification metrics.
`extract_shapes`	Calculate polygon shapes of each element of the confusion matrix.

API¶

digger.compare.confusion.calculate_metrics(true_positive: float, true_negative: float, false_positive: float, false_negative: float, yamlout: str = 'confusion_metrics.yaml') → dict¶

Calculate binary classification metrics.

See the confusion matrix Wikipedia page for additional detail regarding many of the provided classification metrics. Note that many of these metrics have multiple names. For simplicity, only one name is written here.

\(\alpha_T\), \(\beta_T\), \(\gamma_T\), and \(\Omega_T\) are from Schraml et al. (2016).

Inputs:

true_positivefloat: Number of true positives (TP)
true_negativefloat: Number of true negatives (TN)
false_positivefloat: Number of false positives (FP)
false_negativefloat: Number of false negatives (FN)
yamloutstr: Name of a yaml file to dump the calculated metrics.

Outputs:

outdict

Dictionary with the following columns

Table 7 Returned dictionary key value definition.¶
Key	Description of value	Equation
ppv	Positive predictive value	\(\frac{TP}{TP + FP}\)
tpr	True positive rate	\(\frac{TP}{TP + FN}\)
tnr	True negative rate	\(\frac{TN}{TN + FP}\)
npv	Negative predictive value	\(\frac{TN}{TN + FN}\)
fnr	False negative rate	\(\frac{FN}{FN + TP}\)
fpr	False positive rate	\(\frac{FP}{FP + TN}\)
fdr	False discovery rate	\(\frac{FP}{FP + TP}\)
for	False omission rate	\(\frac{FN}{FN + TN}\)
f1	\(F_1\) score	\(\frac{2TP}{2TP + FP + FN}\)
ts	threat score	\(\frac{TP}{TP + FN + FP}\)
accuracy	Accuracy	\(\frac{TP + TN}{TP + FP + TN + FN}\)
alphaT	\(\alpha_T\)	\(\frac{TP}{TP + FN + FP}\)
betaT	\(\beta_T\)	\(\frac{FN}{TP + FN + FP}\)
gammaT	\(\gamma_T\)	\(\frac{FP}{TP + FN + FP}\)
OmegaT	\(\Omega_T\)	\(\frac{TP - FP - FN}{TP + FN + FP}\)

References

Schraml, K., Thomschitz, B., McArdell, B. W., Graf, C., & Kaitna, R. (2015). Modeling debris-flow runout patterns on two alpine fans with different dynamic simulation models. Natural Hazards and Earth System Sciences, 15(7), 1483–1492. https://doi.org/10.5194/nhess-15-1483-2015

digger.compare.confusion.extract_shapes(truth_fn: str, simulation_fn: str, region_fn: str | None = None, make_plot: bool = True, topo_file: str = None, initial_source: str = None) → dict¶

Calculate polygon shapes of each element of the confusion matrix.