[Python] dtreeviz python package is showing a split as being at 0 instead of the actual number

The dtreeviz python package is showing a split at .216 as "0.000" as seen in this screenshot:

The code is:

self.clf = DecisionTreeClassifier(random_state=1234, max_depth=self.treeDepth)
viz_model = dtreeviz.model(self.clf,
X_train=self.features, y_train=self.target,
feature_names=self.featureColumns,
target_name=self.targetColumn, class_names=["short", "long"])
v = viz_model.view(scale=2, precision=4)
v.save(path)


The textual representation shows the splits correctly:

Code:

text_representation = tree.export_text(self.clf, decimals=3)


Output:

|--- feature_1 <= 0.216
| |--- feature_1 <= 0.212
| | |--- feature_1 <= 0.208
| | | |--- class: 0
| | |--- feature_1 > 0.208
| | | |--- class: 1
| |--- feature_1 > 0.212
| | |--- class: 0
|--- feature_1 > 0.216
| |--- feature_1 <= 0.401
| | |--- feature_1 <= 0.379
| | | |--- class: 1
| | |--- feature_1 > 0.379
| | | |--- class: 1
| |--- feature_1 > 0.401
| | |--- feature_1 <= 0.430
| | | |--- class: 0
| | |--- feature_1 > 0.430
| | | |--- class: 1

Continue reading...