CODEWITHSUNDEEP
pythonpandasexportscikit-learn
progster
progster

Reputation: 937

exporting Python model results

Hi I’ve launched a random forest over a dataset imported as df. Now I would like to export both results (0-1 prediction) and predicted probabilities ( a two dimensions array) and match them to my dataset df. Is that possible? Until now I figured out how to export in a separate way to csv. And yes, I am not a pandas expert yet. Any hint? 

# Import the `RandomForestClassifier`
from sklearn.ensemble import RandomForestClassifier


# Create the target and features numpy arrays: 

target = df["target"].values


features =df[["var1",
"var2","var3","var4","var5"]]


features_forest = features

# Building and fitting my_forest
forest = RandomForestClassifier(max_depth = 10, min_samples_split=2, n_estimators = 200, random_state = 1)
my_forest = forest.fit(features_forest, target)

# Print the score of the fitted random forest
print(my_forest.score(features_forest, target))


print(my_forest.feature_importances_)


results = my_forest.predict(features)

print(results)

predicted_probs = forest.predict_proba(features)

#predicted_probs = my_forest.predict_proba(features)

print(predicted_probs)

id_test = df['ID_CONTACT']


pd.DataFrame({"id": id_test, "relevance": results, "probs": predicted_probs }).to_csv('C:\Users\me\Desktop\python\data\submission.csv',index=False)


pd.DataFrame(predicted_probs).to_csv('C:\Users\me\Desktop\python\data\submission_2.csv',index=False)

Upvotes: 0

Views: 1220

Answers (1)

Stefan
Stefan

Reputation: 42885

You should be able to 

df['results] = results
df = pd.concat([df, pd.DataFrame(predicted_probs, columns=['Col_1', 'Col_2'])], axis=1)

Upvotes: 1

Related Questions