mozilla#3 data-split space mapped

dev/Addi-11/classifiers.py

@@ -6,6 +6,8 @@
 from sklearn.tree import DecisionTreeClassifier
 from sklearn.ensemble import RandomForestClassifier
 from sklearn.naive_bayes import GaussianNB
+from sklearn.metrics import plot_precision_recall_curve, confusion_matrix, plot_confusion_matrix
+import matplotlib.pyplot as plt
 
 class Classifier:
 
@@ -39,12 +41,17 @@ def Gaussian(self, X_train,y_train):
         classifier.fit(X_train, y_train)
         return classifier
 
+    """ This function is to evaluate classifier's performancen"""
     def evaluation(self, classifier, X_test, y_test):
-        accuracy, precision, recall, f_score = evaluate(classifier, X_test, y_test)
+        accuracy, precision, recall, f_score , y_score = evaluate(classifier, X_test, y_test)
         print("Accuracy : ",accuracy)
         print("Precision: ", precision)
         print("Recall: ", recall)
         print("F1 score : ",f_score)
+        print("Precision vs Recall Curve")
+        disp = plot_precision_recall_curve(classifier,X_test, y_test)
+        print("Confusion Matrix")
+        disp = plot_confusion_matrix(classifier,X_test, y_test)
 
 
 

dev/Addi-11/data_split_examine.py

@@ -0,0 +1,34 @@
+""" This file compares various evaluation metrics for different data splits """
+
+import numpy as np
+from dataloader import train_test_split_data
+from evaluation import evaluate
+from classifiers import Classifier
+import pandas as pd
+from IPython.display import HTML
+from pylab import *
+
+""" for now KNeighbors will be used as it gave the highest accuracy """
+model = Classifier()
+
+test_sizes = np.arange(0.0001,1,0.05)
+columns = ['Training data','Testing Data','Accuracy', 'Precision', 'Recall', 'F1_score']
+df = pd.DataFrame(columns = columns)
+
+def data_split_examine():
+    for index in range(len(test_sizes)):
+        X_train, X_test, y_train, y_test = train_test_split_data(test_sizes[index])
+        classifier = model.KNeighbors(X_train, y_train)
+        accuracy, precision, recall, f_score, y = evaluate(classifier, X_test, y_test)
+        train = round((1-test_sizes[index])*100)
+        test = round(test_sizes[index]*100)
+        df.loc[index+1] = [train, test, accuracy, precision, recall, f_score]
+
+    display(df)
+
+def visualise_split():
+    fig,axes = plt.subplots()
+    axes.set_xlabel("Accuracy")
+    axes.set_ylabel("Test Data Size")
+    axes.set_title("Relation btw accuracy and test data size")
+    disp = axes.plot(test_sizes, df.Accuracy)

dev/Addi-11/dataloader.py

@@ -5,6 +5,7 @@
 import pandas as pd
 import numpy as np
 from sklearn.model_selection import train_test_split
+import seaborn as sns
 
 """ read data file """
 address = "../../datasets/eeg.csv"
@@ -21,6 +22,10 @@ def data_info():
     display(df.info()) # info of each column of dataset
     print("\nChecking for null values: \n")
     print(df.isnull().sum()) # sum of all null values in a dataset for preprocessing    
+""" Function specific to given dataset"""
+def data_visuals():
+    df.Class.unique()
+    sns.countplot(df.Class)
 
 """ splitting the data """
 def train_test_split_data(test_size):