added more algorithms

CoopClassifier.py

@@ -2,8 +2,13 @@ from sklearn.preprocessing import MultiLabelBinarizer
 from sklearn.model_selection import train_test_split
 from sklearn.feature_extraction.text import CountVectorizer
 from sklearn.multiclass import OneVsRestClassifier
-from sklearn.svm import SVC, LinearSVC
 from sklearn import metrics
+
+from sklearn.svm import SVC, LinearSVC
+from sklearn.linear_model import LogisticRegression, SGDClassifier, Perceptron, PassiveAggressiveClassifier
+from sklearn.ensemble import GradientBoostingClassifier, RandomForestClassifier
+from sklearn.tree import DecisionTreeClassifier
+
 import json
 import numpy
 import warnings
@@ -35,14 +40,21 @@ X = [' '.join(inner_list) for inner_list in
 
 
 vect = CountVectorizer()
-clf = OneVsRestClassifier(SVC())
+clf = RandomForestClassifier()
+# clf = DecisionTreeClassifier()
+# clf = OneVsRestClassifier(PassiveAggressiveClassifier())
+# clf = OneVsRestClassifier(Perceptron())
+# clf = OneVsRestClassifier(LogisticRegression(multi_class='ovr'))
+# clf = OneVsRestClassifier(GradientBoostingClassifier())
+# clf = OneVsRestClassifier(SGDClassifier())
+# clf = OneVsRestClassifier(SVC())
 # clf = OneVsRestClassifier(LinearSVC()) # For larger data sets
 mlb = MultiLabelBinarizer()  # Creates a binary matrix with labels as columns, sorted by name. Samples as rows.
 
 # OBS! Om en ingrediens är flera ord "färska örter" blir "färska" en ingrediens. Ordna (kanske).
 y_binary_matrix = mlb.fit_transform(Y)  # kolumner, labels, är ordnade efter bokstavsordning, enligt mlb.classes_
 x_binary_matrix = vect.fit_transform(X)  # kolumner är ordnade efter bokstavsordning, enligt vectorizer.get_feature_names()
-train_x, test_x, train_y, test_y = train_test_split(x_binary_matrix, y_binary_matrix, test_size=0.1)
+train_x, test_x, train_y, test_y = train_test_split(x_binary_matrix, y_binary_matrix, test_size=0.15)
 # Lower test size => larger training size => better predict metrics
 
 clf.fit(train_x, train_y)
@@ -52,7 +64,7 @@ predictions = clf.predict(test_x)
 customTestSample = vect.transform(['havregryn mjölk salt hallon', 'kycklingfärs ströbröd mjölk ägg persilja citron svartpeppar smör salt potatis', 'smör färskost ägg mandel hasselnöt råsocker vetemjöl'])
 print(mlb.inverse_transform(clf.predict(customTestSample)))
 
-result_metrics = metrics.classification_report(test_y, predictions, target_names=mlb.classes_)
+result_metrics = metrics.classification_report(test_y, predictions, target_names=mlb.classes_, zero_division=True)
 print(result_metrics)
 
 
@@ -74,3 +86,4 @@ print(result_metrics)
 "Måltid|Fest" får dålig prediction för att det finns 4 recept med det, dock många fler med Tillfällen|fest
 Veckans middagar får dålig prediction för att..?=?
 """
+