Model.classification_report¶

In [ ]:

Model.classification_report(cutoff = [],
                            labels: list = [])

Computes a classification report using multiple metrics to evaluate the model (AUC, accuracy, PRC AUC, F1...). In case of multiclass classification, it will consider each category as positive and switch to the next one during the computation.

Parameters¶

Name	Type	Optional	Description
cutoff	float / list	✓	Cutoff for which the tested category will be accepted as prediction. In case of multiclass classification, each tested category becomes the positives and the others are merged into the negatives. The list will represent the classes threshold. If it is empty or invalid, the best cutoff will be used.
labels	list	✓	List of the different labels to be used during the computation.

Returns¶

tablesample : An object containing the result. For more information, see utilities.tablesample.

Example¶

In [49]:

# Multiclass Classification
from verticapy.learn.ensemble import RandomForestClassifier
model = RandomForestClassifier(name = "public.RF_iris",
                               n_estimators = 20,
                               max_features = "auto",
                               max_leaf_nodes = 32, 
                               sample = 0.7,
                               max_depth = 3,
                               min_samples_leaf = 5,
                               min_info_gain = 0.0,
                               nbins = 32)
model.fit("public.iris", ["PetalLengthCm", "PetalWidthCm"], "Species")
# Multiclass Classification: Using a fixed cutoff
model.classification_report(cutoff = 0.33)

	Iris-setosa	Iris-versicolor	Iris-virginica
auc	1.0	0.9970000000000002	0.9972000000000002
prc_auc	1.0	0.9942642287130048	0.9945555192416813
accuracy	1.0	0.9733333333333334	0.96
log_loss	0.0080071224076684	0.0304476426619168	0.0292307724947753
precision	1.0	0.96	0.9074074074074074
recall	1.0	0.96	0.98
f1_score	1.0	0.9698969072164948	0.9647668393782383
mcc	1.0	0.94	0.9133462590326239
informedness	1.0	0.94	0.9299999999999999
markedness	1.0	0.94	0.8969907407407409
csi	1.0	0.9230769230769231	0.8909090909090909
cutoff	0.33	0.33	0.33

Out[49]:

In [50]:

# Multiclass Classification: Using automatic cutoffs
model.classification_report()

	Iris-setosa	Iris-versicolor	Iris-virginica
auc	1.0	0.9970000000000002	0.9972000000000002
prc_auc	1.0	0.9942642287130048	0.9945555192416813
accuracy	1.0	0.9733333333333334	0.9666666666666667
log_loss	0.0080071224076684	0.0304476426619168	0.0292307724947753
precision	1.0	0.96	0.9090909090909091
recall	1.0	0.96	1.0
f1_score	1.0	0.9698969072164948	0.9743589743589743
mcc	1.0	0.94	0.929320377284585
informedness	1.0	0.94	0.95
markedness	1.0	0.94	0.9090909090909092
csi	1.0	0.9230769230769231	0.9090909090909091
cutoff	0.851	0.382	0.284

Out[50]:

In [41]:

# Multiclass Classification: Customized Cutoffs
model.classification_report(cutoff = [0.8, 0.4, 0.2])

	Iris-setosa	Iris-versicolor	Iris-virginica
auc	1.0	0.996	0.9960000000000001
prc_auc	1.0	0.9925238799641174	0.9921025338869698
accuracy	1.0	0.9466666666666667	0.9666666666666667
log_loss	0.00993790055646546	0.0375615053542112	0.0344348556530513
precision	1.0	0.875	0.9090909090909091
recall	1.0	0.98	1.0
f1_score	1.0	0.9543455497382198	0.9743589743589743
mcc	1.0	0.8868873372284499	0.929320377284585
informedness	1.0	0.9100000000000001	0.95
markedness	1.0	0.8643617021276597	0.9090909090909092
csi	1.0	0.8596491228070176	0.9090909090909091
cutoff	0.8	0.4	0.2

Out[41]:

In [42]:

# Multiclass Classification: Choosing the categories
model.classification_report(labels = ["Iris-versicolor", "Iris-virginica"])

	Iris-versicolor	Iris-virginica
auc	0.996	0.9960000000000001
prc_auc	0.9925238799641174	0.9921025338869698
accuracy	0.9533333333333334	0.9666666666666667
log_loss	0.0375615053542112	0.0344348556530513
precision	0.8771929824561403	0.9090909090909091
recall	1.0	1.0
f1_score	0.9637305699481865	0.9743589743589743
mcc	0.9032106474595007	0.929320377284585
informedness	0.9300000000000002	0.95
markedness	0.8771929824561404	0.9090909090909092
csi	0.8771929824561403	0.9090909090909091
cutoff	0.323	0.23

Out[42]: