mlscorer 0.3

No project description provided

ML Scorer

ML Scorer is the solution to your classification scores of ML algorithms.

Installation

pip install mlscorer

Preperation

Make a class mapping dictionary(map_class) using Method1 or Method2

Method 1

Make all the data categorical using following code snippet

map_class = dict(zip(df.classes.astype("category").cat.codes, df.classes))
print(map_class)

output: {1: 'positive', 0: 'negative'}

here, df is the Dataframe and classes is a column which may have class values like

• positive
• negative

[N.B. Don't change "category", it's a datatype]

or

Method 2

Make the Dictionary manually according to your classes

map_class = {
    1: 'positive',
    0: 'negative'
}

Usage

from sklearn.linear_model import LogisticRegression
import mlscorer as ms
classifier = LogisticRegression()
classifier.fit(X_train, y_train)
y_pred = classifier.predict(X_test)
ms.get_score_table(y_test, y_pred, map_class)

Output:

Parameters

y_test : target values of test set

y_pred : predicted target values

map_class : dict : your categoricl class mapping

metrics : list : use one or more evaluation metric from f1, precision, recall or accuracy

eg:

ms.get_score_table(y_test, y_pred, map_class, metrics=['precision', 'recall'])

Output: