How to print classification report in FastText?

Question

I'm using FastText and to evaluate the results of my classification (binary classification) and I would like to print the Clasisfication Score. Actually as output I'm having the Precision and Recall. Here my code:

train_file = 'train.csv'

test_file = 'test.csv' 


print("training model...")
  
    
model = fasttext.train_supervised(input=train_file,
                                lr=1.0, epoch=100,
                                wordNgrams=2, 
                                bucket=200000, 
                                dim=50, 
                                loss='hs')

def print_results(N, p, r):
    print("N\t" + str(N))
    print("P@{}\t{:.3f}".format(1, p))
    print("R@{}\t{:.3f}".format(1, r))
    

result = model.test(test_file)
print_results(*result)

I was reading the documentation here and checking also some tutorials but I did not understand yet how to define the classification report. I think that a loop through the test_file should be defined but in this case, which parameters I have to use in?

Desired output:

            precision    recall  f1-score   support

           0      0.832     0.824     0.828      9093
           1      0.861     0.867     0.864     11399

    accuracy                          0.848     20492
   macro avg      0.846     0.846     0.846     20492
weighted avg      0.848     0.848     0.848     20492

Thanks,

Answer 1

Why don't you deal with metrics library from scikit-learn , maybe is a better choice.

F1 Code example:

f1_score('your_test', 'your_predict', average='macro')

In your case I think should be:

f1_score(test_file, result, average='macro')

Now you can make a table with all scores you want.

You just have to import them, for example:

from sklearn.metrics import  f1_score, precision_score, recall_score, accuracy_score

And you can even add more things, for the example 'confusion matrix':

from sklearn.metrics import confusion_matrix

Code example:

confusion_matrix(test_file, result)