Confusion Matrix is of wrong shape #19756
Description
Describe the bug
Having 100 percent true positive rate gives a 1x1 matrix. This can cause bugs if you are using the confusion matrix in a function to calculate evaluation parameters such as sensitivity, auc, specifity, accuracy.
from sklearn.metrics import confusion_matrix, roc_auc_score
ground_truth, predictions = np.ones((2,1))
cf_matrix = confusion_matrix(ground_truth, predictions)
assert cf_matrix.shape == (2,2), ("The confusion matrix is of shape"+str(cf_matrix.shape))
Expected Results
No error is thrown.
Actual Results
AssertionError Traceback (most recent call last)
<ipython-input-23-9c79b721fccb> in <module>
3 ground_truth, predictions = np.ones((2,1))
4 cf_matrix = confusion_matrix(ground_truth, predictions)
----> 5 assert cf_matrix.shape == (2,2), ("The confusion matrix is of shape"+str(cf_matrix.shape))
AssertionError: The confusion matrix is of shape(1, 1).
Versions
System:
python: 3.7.10 | packaged by conda-forge | (default, Feb 19 2021, 15:59:12) [Clang 11.0.1 ]
executable: /Users/julian/opt/anaconda3/envs/cadd-course/bin/python
machine: Darwin-20.3.0-x86_64-i386-64bit
Python dependencies:
pip: 21.0.1
setuptools: 49.6.0.post20210108
sklearn: 0.24.1
numpy: 1.20.1
scipy: 1.6.1
Cython: None
pandas: 1.2.3
matplotlib: 3.3.4
joblib: 1.0.1
threadpoolctl: 2.1.0
Built with OpenMP: True