CI/test_ci.py

# Utilities
import sys
import torch
sys.modules['torch'] = torch 
import os
os.environ["CUDA_VISIBLE_DEVICES"] = ""
from src.modello import Modello 
from src.dataset import LoadDataset
from sklearn.metrics import accuracy_score

class TestClass :
    """
    unit test sul modello per pipeline CI
    """ 
    
    def test_trivial_output(self) :
        # Controllo del funzionamento del modello con frasi banali
        model = Modello()
        assert model.predict("neutral")[0]=="neutral" and model.predict("awesome")[0]=="positive" and model.predict("terrible")[0]=="negative"
    
    def test_accuracy(self) :
        # Controllo che l'accuracy sia almeno 0.5
        model = Modello()
        ld = LoadDataset()
        X = ld.X
        y = ld.y
        y_pred = model.predict(X)
        assert accuracy_score(y, y_pred)>=0.5