name error:name 'clf' is not defined

saiankith · Mar-31-2020, 07:23 AM

Quote:I am working on a project called heart disease predictor.When i click the button predict heart disease.The below error is coming.But clf is defined global.Help me out asap.
Thank you.

        
          
          
              
              import tkinter #importing tkinter library for GUI creation
from tkinter import *
from PIL import Image, ImageTk
import tkinter.messagebox
from tkinter import filedialog
#import mysql.connector
#from keras.models import Sequential
#from keras.layers import Dense
import pandas as pnd  # importing pandas data analysis toolkit
import numpy as np    # importing numpy library for array operations
from time import time # importing time library for time calculations
from sklearn.model_selection import train_test_split # importing module model_classification from scikit-learn library
print("hello")
    
header_row = ['age', 'sex', 'pain', 'BP', 'chol', 'fbs', 'ecg', 'maxhr','eiang', 'eist', 'slope', 'vessels', 'thal',
              'diagnosis']     # Declaring the header row for getting data from the dataset files
 
 
# filter to only those diagnosed with heart disease
def cardiac():
    global master
    master = Tk()         # Defining the Tkinter widget
    master.wm_title("Heart Disease Prediction")
    master.geometry('1500x600')
    image = Image.open('c3.jpg')
    image = image.resize((1500, 600))
    photo_image = ImageTk.PhotoImage(image)
    label = Label(master, image = photo_image)
    label.place(x=0,y=0)
    import sklearn        # Importing scikit-learn functions
 
    #Lab=Label(master,text=" Automatic Heart Disease Detection ")    # Adding Label to the Tkinter widget
    #Lab.place(x=600,y=50)                           # Packing the label data to the tkinter widget in user defined rows and columns
                                         # Changing dimensions of the Label
 
    #Lab=Label(master,text="")
    #Lab.grid(row=2,column=5,columnspan=2)
 
    #Lab1=Label(master,text="Classification Report")
    #Lab1.place(x=170,y=330)
    #Lab2=Label(master,text="Confusion Matrix")
    #Lab2.place(x=930,y=330)    
    T = Text(master, height=6, width=40,font=("bold",10),highlightthickness=2,bg="white",relief=SUNKEN)                            # Declaring Text Widget for Result Displaying
    T.place(x=55,y=350)
    T1 = Text(master, height=6, width=35,font=("bold",10),highlightthickness=2,bg="white",relief=SUNKEN)
    T1.place(x=1000,y=350)
 
    var = StringVar(master)
    var.set("Select Dataset") # initial value
 
    option = OptionMenu(master, var, "Cleveland", "Hungarian", "VA", "all") # Declaring the OptionMenu (Drop-Down list) widget
    option.config(bg = "violet")
    option.config(fg = "black")
    option.config(font=('algerian',10,'bold'))
    option.config(width=12)
    option.place(x=500,y=80)
 
    '''field1="Age"                                                            # Defining the field names which user has to input for heart disease detection
    field2="Sex"
    field3="Pain"
    field4="BP"
    field5="Chol"
    field6="FBS"
    field7="ECG"
    field8="Maxhr"
    field9="Eiang"
    field10="Eist"
    field11="Slope" 
    field12="Vessels"
    field13="Thal"'''
 
 
    '''L1=Label(master,text=field1)
    L1.grid(row = 4, column = 0, sticky='nsew')
    L1.configure(width=14)
    L2=Label(master,text=field2)
    L2.grid(row = 4, column = 1, sticky='nsew')
    L2.configure(width=14)
    L3=Label(master,text=field3)
    L3.grid(row = 4, column = 2, sticky='nsew')
    L3.configure(width=14)
    L4=Label(master,text=field4)
    L4.grid(row = 4, column = 3, sticky='nsew')
    L4.configure(width=14)
    L5=Label(master,text=field5)
    L5.grid(row = 4, column = 4, sticky='nsew')
    L5.configure(width=14)
    L6=Label(master,text=field6, )
    L6.grid(row = 4, column = 5, sticky='nsew')
    L6.configure(width=14)
    L7=Label(master,text=field7)
    L7.grid(row = 4, column = 6, sticky='nsew')
    L7.configure(width=14)
    L8=Label(master,text=field8)
    L8.grid(row = 4, column = 7, sticky='nsew')
    L8.configure(width=14)
    L9=Label(master,text=field9)
    L9.grid(row = 4, column = 8, sticky='nsew')
    L9.configure(width=14)
    L10=Label(master,text=field10)
    L10.grid(row = 4, column = 9, sticky='nsew')
    L10.configure(width=14)
    L11=Label(master,text=field11)
    L11.grid(row = 4, column = 10, sticky='nsew')
    L11.configure(width=14)
    L12=Label(master,text=field12)
    L12.grid(row = 4, column = 11, sticky='nsew')
    L12.configure(width=14)
    L13=Label(master,text=field13)
    L13.grid(row = 4, column = 12, sticky='nsew')
    L13.configure(width=14)'''
 
 
    E1=Entry(master,width=8,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E1.place(x=0, y=220)
 
    E2=Entry(master,width=8,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E2.place(x=90, y=220)
 
    E3=Entry(master,width=8,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E3.place(x=180, y=220)
 
    E4=Entry(master,width=8,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E4.place(x=280, y=220)
 
    E5=Entry(master,width=8,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E5.place(x=370, y=220)
 
    E6=Entry(master,width=8,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E6.place(x=470, y=220)
 
    E7=Entry(master,width=8,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E7.place(x=550, y=220)
 
    E8=Entry(master,width=8,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E8.place(x=650, y=220)
 
    E9=Entry(master,width=10,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E9.place(x=770, y=220)
 
    E10=Entry(master,width=10,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E10.place(x=880, y=220)
 
    E11=Entry(master,width=10,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E11.place(x=980, y=220)
 
    E12=Entry(master,width=10,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E12.place(x=1100, y=220)
 
    E13=Entry(master,width=10,font=("bold",10),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E13.place(x=1220, y=220)
 
    lb1 = Label(master, text="patient",font=('algerian',15,'bold'),fg="BLACK",anchor='w')
    lb1.place(x=0, y=150)
 
    E0=Entry(master,width=10,font=("bold",15),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    E0.place(x=120, y=150)
 
 
 
    '''Labx=Label(master,text="")
    Labx.grid(row=21,column=4,columnspan=4)
    Labx.visible=False'''
 
    #T3 = Text(master, height=2, width=30)                                                       # Declaring Text Widget for Displaying Prediction
    #T3.grid(row=23,column=4, columnspan=4, sticky= 'nsew')
 
    def train_classifier(x_train,x_test,y_train,y_test,string):                         # Declaring the function for training classifiers and classification analysis
        global clf                                                                      # Declaring clf as a Global Variable for using throughot the code
        global outclass
        global a1,a2,a3
        from sklearn.metrics import classification_report
        from sklearn.metrics import confusion_matrix
        T.delete(1.0,END)                                                               # Deleting the text in the Text Widget
        T1.delete(1.0,END)
        if string=="CNN":
            
            from keras.models import Sequential
            from keras.layers import Dense
            import matplotlib.pyplot as plt1
             
            #import pandas as pnd  # importing pandas data analysis toolkit
            #import numpy as np 
            t1= time()
              
            clf = Sequential() #initial creation
            clf.add(Dense(13, input_dim=13, init='uniform', activation='relu')) #first hidden layer
            clf.add(Dense(10, init='uniform', activation='relu'))
            #model.add(Dense(8, init='uniform', activation='relu'))
            #model.add(Dense(6, init='uniform', activation='relu'))
            clf.add(Dense(1, init='uniform', activation='sigmoid')) #output layer
 
            # compile the model
            clf.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
 
            # fitting data to model
            clf.fit(x_train, y_train, validation_data=(x_test, y_test), epochs=200, batch_size=5, verbose=0)
 
            # evaluate the model
            scores = clf.evaluate(x_test, y_test)
            #printing accuracy
            print("Accuracy: %.2f%%" % (scores[1]*100))
            #p1=model.predict(Xv)
            #print(p1)
            a1=(scores[1]*100)
 
            T.insert(END, "Accuracy")
            T.insert(END, "  ")
            T.insert(END,(scores[1]*100) )
            y_pred = clf.predict(x_test)
            print(y_pred)
            l1=list()
            for i in y_pred:
                if i<.49:
                    l1.append(0)
                else:
                    l1.append(1)
            print(l1)
 
            count1=y_test.count()
            l2=list(range(1,count1+1,1))
             
 
            print(count1)
            print(y_test.count())
            #print(count(l1))
 
            colormap=np.array(['lime','red'])
            plt1.subplot(1,2,1)
            plt1.scatter(l2,y_test,c=colormap[y_test])
            plt1.suptitle("CNN Algorihm")
            plt1.xlabel("X values")
            plt1.ylabel("y tests values")
            plt1.title("true Values")
 
            plt1.subplot(1,2,2)
            plt1.scatter(l2,l1,c=colormap[l1])
            plt1.suptitle("CNN Algorihm")
            plt1.xlabel("X values")
            plt1.ylabel("Predictions")
            plt1.title("Predictions")
            plt1.show()
             
            # Predict Results for Test Data
            #title = "Learning Curves (SVM)"
            #print(y_train)
            #geterror(x_train,y_train,clf,title);
            t= time()-t1
            T.insert(END, "Accuracy")
            T.insert(END, "  ")
            T.insert(END,(scores[1]*100) )
            print("Training Complete")
         
        elif string=="Naive Bayes":
            from sklearn.naive_bayes import GaussianNB
            import matplotlib.pyplot as plt2
             
            t2=time()
            clf = GaussianNB()                                                          # Initializing the Naive Bayes Classifier
            clf.partial_fit(x_train, y_train, np.unique(y_train))                       # Fitting the classifier to the training and testing the Naive Bayes Classifier
            y_pred = clf.predict(x_test)
            print(y_pred)
            t = time() - t2
            title = "Learning Curves (Naive Bayes)"
 
             
            #geterror(x_train,y_train,clf,title);
            classre=classification_report(y_test,y_pred)                       # Generating Classification Report
            T.insert(END,classre[1:5]+classre[1:32]+classre[1:13]+classre[60:90]+classre[1:11]+classre[1:2]+classre[115:140]+classre[1:7]+classre[161:195]) # Printing Precision and Recall Results 
            print(classre)
            confmat=confusion_matrix(y_test,y_pred)                                        # Calculating the Confusion Matrix for the classification
            T1.insert(END, confmat)
            T.insert(END, classre[1:9])
            T.insert(END, "Accuracy")
            T.insert(END, classre[1:5])
            T.insert(END, int(float((y_test==y_pred).sum())/len(y_test.T)*100))
            T.insert(END, "%")
            T.insert(END, classre[1:10]+classre[1:10])
            T.insert(END, "Class. Time")
            T.insert(END, classre[1:8])
            #T.insert(END, t[0:4]+" sec")
            a2=int(float((y_test==y_pred).sum())/len(y_test.T)*100)
         
            
 
            count2=y_test.count()
            l3=list(range(1,count2+1,1))
             
 
            print(count2)
            print(y_test.count())
            #print(count(l1))
 
            colormap=np.array(['lime','red'])
            plt2.subplot(1,2,1)
            plt2.scatter(l3,y_test,c=colormap[y_test])
            plt2.suptitle("Naviee Bayesian Algorihm")
            plt2.xlabel("X values")
            plt2.ylabel("y tests values")
            plt2.title("true Values")
 
            plt2.subplot(1,2,2)
            plt2.scatter(l3,y_pred,c=colormap[y_pred])
            plt2.suptitle("Naviee Bayesian Algorihm")
            plt2.xlabel("X values")
            plt2.ylabel("Predictions")
            plt2.title("Predictions")
            plt2.show()
            print("Training Complete")
             
         
         
        elif string=="K-Nearesr Neighbour":
            from sklearn.neighbors import KNeighborsClassifier
            import matplotlib.pyplot as plt3
            clf = KNeighborsClassifier(n_neighbors=5)
            clf.fit(x_train, y_train)
            y_pred = clf.predict(x_test)
            print(y_pred)
            print(y_test)
            classre=classification_report(y_test,y_pred)                       # Generating Classification Report
            T.insert(END,classre[1:5]+classre[1:32]+classre[1:13]+classre[60:90]+classre[1:11]+classre[1:2]+classre[115:140]+classre[1:7]+classre[161:195]) # Printing Precision and Recall Results 
            print(classre)
            confmat=confusion_matrix(y_test,y_pred)                                        # Calculating the Confusion Matrix for the classification
            T1.insert(END, confmat)
            T.insert(END, classre[1:9])
            T.insert(END, "Accuracy")
            T.insert(END, classre[1:5])
            T.insert(END, int(float((y_test==y_pred).sum())/len(y_test.T)*100))
            T.insert(END, "%")
            T.insert(END, classre[1:10]+classre[1:10])
            #T.insert(END, "Class. Time")
            #T.insert(END, classre[1:8])
            #T.insert(END, t[0:4]+" sec")
            a3=int(float((y_test==y_pred).sum())/len(y_test.T)*100)
     
            count3=y_test.count()
            l4=list(range(1,count3+1,1))
             
 
            print(count3)
            print(a1)
            print(a2)
            print(a3)
            #plot_bar_x(a1,a2,a3);
            print(y_test.count())
            #print(count(l1))
 
            colormap=np.array(['lime','red'])
            plt3.subplot(1,2,1)
            plt3.scatter(l4,y_test,c=colormap[y_test])
            plt3.suptitle("K Nearest Neighbour Algorihm")
            plt3.xlabel("X values")
            plt3.ylabel("Y tests values")
            plt3.title("True Values")
 
            plt3.subplot(1,2,2)
            plt3.scatter(l4,y_pred,c=colormap[y_pred])
            plt3.suptitle("K Nearest Neighbour Algorihm")
            plt3.xlabel("X values")
            plt3.ylabel("Predictions")
            plt3.title("Predictions")
            plt3.show()
            #print(a1)
            #print(a2)
            #print(a3)
            plot_bar_x(a1,a2,a3)
            print("Training Complete")
            
 
        '''elif string=="Logistic Regression":
            from sklearn.linear_model import LogisticRegression
            t3=time()
            clf=LogisticRegression(penalty='l2', dual=False, tol=0.0001, C=1.0, fit_intercept=True, intercept_scaling=1, solver='liblinear', max_iter=100, verbose=0, warm_start=False, n_jobs=1)   # Initializing the Logistic Regression Classifier
            clf.fit(x_train,y_train)
            y_pred = clf.predict(x_test)
            print(y_pred)
            t = time() - t3
            title = "Learning Curves (Logistic Regression)"
           # geterror(x_train,y_train,clf,title);
            print("Training Complete")
 
            t=str(t)'''
        '''print(y_test)
        classre=classification_report(y_test,y_pred)                       # Generating Classification Report
        T.insert(END,classre[1:5]+classre[1:32]+classre[1:13]+classre[60:90]+classre[1:11]+classre[1:2]+classre[115:140]+classre[1:7]+classre[161:195]) # Printing Precision and Recall Results 
        print(classre)
        confmat=confusion_matrix(y_test,y_pred)                                        # Calculating the Confusion Matrix for the classification
        T1.insert(END, confmat)
        T.insert(END, classre[1:9])
        T.insert(END, "Accuracy")
        T.insert(END, classre[1:5])
        T.insert(END, int(float((y_test==y_pred).sum())/len(y_test.T)*100))
        T.insert(END, "%")
        T.insert(END, classre[1:10]+classre[1:10])
        T.insert(END, "Class. Time")
        T.insert(END, classre[1:8])
        T.insert(END, t[0:4]+" sec")
        import matplotlib.pyplot as plt'''
         
    def plot_bar_x(a1,a2,a3):
        import matplotlib.pyplot as pltu
        import numpy as np
        print(a1)
        print(a2)
        print(a3)
        label = ['CNN', 'NAIVE BAYESIAN', 'KNN']
        no_movies = [a1,a2,a3]
  
    # this is for plotting purpose
        index = np.arange(len(label))
        pltu.bar(index, no_movies)
        pltu.xlabel('Algorithms', fontsize=15)
        pltu.ylabel('Accuracies', fontsize=15)
        pltu.xticks(index, label, fontsize=15, rotation=30)
        pltu.title('Comparison of different algorithms used')
        pltu.show()
 
    def process_dataset(string):
 
         
     
        if string=="Cleveland":
         
            heart = pnd.read_csv('processed.cleveland.data', names=header_row)          # Reading the dataset file in .data format using Pandas library function read_csv()
            print("Unprocessed Cleveland Dataset")
            print("************************************************************************")
            print(heart.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
 
            import numpy as np
            has_hd_check = heart['diagnosis'] > 0                                                           # Getting the indices of individuals having heart disease
            has_hd_patients = heart[has_hd_check]
            heart['vessels'] = heart['vessels'].apply(lambda vessels: 0.0 if vessels == "?" else vessels)   # Replacing the unknown values in the dataset with float
            heart['vessels'] = heart['vessels'].astype(float)
            heart['thal'] = heart['thal'].apply(lambda thal: 0.0 if thal == "?" else thal)
            heart['thal'] = heart['thal'].astype(float)
            heart['diag_int'] = has_hd_check.astype(int)
 
            ind1 = np.where((heart['diagnosis'] == 1)|(heart['diagnosis'] ==2));
            ind2 = np.where((heart['diagnosis'] == 3)|(heart['diagnosis'] ==4));
 
            temp = heart['diagnosis'];
            temp.ix[ ind1 ] = 1;
            temp.ix[ ind2 ] = 2;
            heart['diagnosis'] = temp;
 
            global x_train
            global y_train
            global x_test
            global y_test
            x_train, x_test, y_train, y_test = train_test_split(heart.loc[:, 'age':'thal'], heart.loc[:, 'diagnosis'],   # Splitting the processed data into training data and testing data
                                                        test_size=0.20, random_state=42)                            # test_size = percent of data used for testing,
                                                                                                                    # random_state = for initializing the random number generator
 
            print("Processed Cleveland Dataset")
            print("************************************************************************")
            print(heart.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
         
 
        elif string=="VA":
 
            import numpy as np
 
 
         
            heart_va = pnd.read_csv('processed.va.data', names=header_row)
            print("Unprocessed VA Dataset")
            print("************************************************************************")
            print(heart_va.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
 
            has_hd_check = heart_va['diagnosis'] > 0
 
         
         
            heart_va['diag_int'] = has_hd_check.astype(int) 
            heart_va = heart_va.replace(to_replace='?', value=0.0)
            heart_va['diag_int'] = has_hd_check.astype(int)
         
            ind1 = np.where((heart_va['diagnosis'] == 1)|(heart_va['diagnosis'] ==2));
            ind2 = np.where((heart_va['diagnosis'] == 3)|(heart_va['diagnosis'] ==4));
 
            temp = heart_va['diagnosis'];
            temp.ix[ ind1 ] = 1;
            temp.ix[ ind2 ] = 2;
 
            heart_va['diagnosis'] = temp;
         
            print("Processed VA Dataset")
            print("************************************************************************")
            print(heart_va.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
          
            x_train, x_test, y_train, y_test = train_test_split(heart_va.loc[:, 'age':'thal'], heart_va.loc[:, 'diagnosis'],
                                                        test_size=0.30, random_state=42)
         
        elif string=="Hungarian":
            import numpy as np
            heart_hu = pnd.read_csv('processed.hungarian.data', names=header_row)
            print("Unprocessed Hungarian Dataset")
            print("************************************************************************")
            print(heart_hu.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
 
            has_hd_check = heart_hu['diagnosis'] > 0
            heart_hu['diag_int'] = has_hd_check.astype(int)
            heart_hu = heart_hu.replace(to_replace='?', value=0.0)
 
            ind1 = np.where((heart_hu['diagnosis'] == 1)|(heart_hu['diagnosis'] ==2));
            ind2 = np.where((heart_hu['diagnosis'] == 3)|(heart_hu['diagnosis'] ==4));
 
            temp = heart_hu['diagnosis'];
            temp.ix[ ind1 ] = 1;
            temp.ix[ ind2 ] = 2;
            heart_hu['diagnosis'] = temp;
 
            print("Processed Hungarian Dataset")
            print("************************************************************************")
            print(heart_hu.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
            heart_hu['diag_int'] = has_hd_check.astype(int)
 
         
            x_train, x_test, y_train, y_test = train_test_split(heart_hu.loc[:, 'age':'thal'], heart_hu.loc[:, 'diagnosis'],
                                                        test_size=0.30, random_state=42)
 
        elif string=="all":
            import numpy as np
            heart_cl = pnd.read_csv('processed.cleveland.data', names=header_row)
            print("Unprocessed Cleveland Dataset")
            print("************************************************************************")
            print(heart_cl.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
            has_hd_check = heart_cl['diagnosis'] > 0
            has_hd_patients = heart_cl[has_hd_check]
            heart_cl['diag_int'] = has_hd_check.astype(int)
            heart_cl['vessels'] = heart_cl['vessels'].apply(lambda vessels: 0.0 if vessels == "?" else vessels)
            heart_cl['vessels'] = heart_cl['vessels'].astype(float)
            heart_cl['thal'] = heart_cl['thal'].apply(lambda thal: 0.0 if thal == "?" else thal)
            heart_cl['thal'] = heart_cl['thal'].astype(float)
 
            ind1 = np.where((heart_cl['diagnosis'] == 1)|(heart_cl['diagnosis'] ==2));
            ind2 = np.where((heart_cl['diagnosis'] == 3)|(heart_cl['diagnosis'] ==4));
 
            temp = heart_cl['diagnosis'];
            temp.ix[ ind1 ] = 1;
            temp.ix[ ind2 ] = 2;
            heart_cl['diagnosis'] = temp;
 
            heart_va = pnd.read_csv('processed.va.data', names=header_row)
            print("Unprocessed VA Dataset")
            print("************************************************************************")
            print(heart_va.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
 
            has_hd_check = heart_va['diagnosis'] > 0
            heart_va['diag_int'] = has_hd_check.astype(int)
            heart_va = heart_va.replace(to_replace='?', value=0.0)
 
            ind1 = np.where((heart_va['diagnosis'] == 1)|(heart_va['diagnosis'] ==2));
            ind2 = np.where((heart_va['diagnosis'] == 3)|(heart_va['diagnosis'] ==4));
 
            temp = heart_va['diagnosis'];
            temp.ix[ ind1 ] = 1;
            temp.ix[ ind2 ] = 2;
            heart_va['diagnosis'] = temp;
 
            print("Processed VA Dataset")
            print("************************************************************************")
            print(heart_va.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
 
            heart_hu = pnd.read_csv('processed.hungarian.data', names=header_row)
            print("Unprocessed Hungarian Dataset")
            print("************************************************************************")
            print(heart_hu.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
 
            has_hd_check = heart_hu['diagnosis'] > 0
            heart_hu['diag_int'] = has_hd_check.astype(int)
            heart_hu = heart_hu.replace(to_replace='?', value=0.0)
 
            ind1 = np.where((heart_hu['diagnosis'] == 1)|(heart_hu['diagnosis'] ==2));
            ind2 = np.where((heart_hu['diagnosis'] == 3)|(heart_hu['diagnosis'] ==4));
 
            temp = heart_hu['diagnosis'];
            temp.ix[ ind1 ] = 1;
            temp.ix[ ind2 ] = 2;
            heart_hu['diagnosis'] = temp;
 
            print("Processed Hungarian Dataset")
            print("************************************************************************")
            print(heart_hu.loc[:, 'age':'diagnosis'])
            print("************************************************************************")
 
            x_train1, x_test1, y_train1, y_test1 = train_test_split(heart_cl.loc[:, 'age':'thal'], heart_cl.loc[:, 'diagnosis'],
                                                        test_size=0.30, random_state=42)
            x_train2, x_test2, y_train2, y_test2 = train_test_split(heart_va.loc[:, 'age':'thal'], heart_va.loc[:, 'diagnosis'],
                                                        test_size=0.30, random_state=42)
            x_train3, x_test3, y_train3, y_test3 = train_test_split(heart_hu.loc[:, 'age':'thal'], heart_hu.loc[:, 'diagnosis'],
                                                        test_size=0.30, random_state=42)
 
            # Combining the dataset for Cleveland, VA and Hungarian Dataset
            x_train4= x_train1.append(x_train2);
            x_train = x_train4.append(x_train3);
         
            y_train4 = y_train1.append(y_train2);
            y_train = y_train4.append(y_train3);
 
            x_test4 = x_test1.append(x_test2);
            x_test = x_test4.append(x_test3)
 
            y_test4 = y_test1.append(y_test2);
            y_test = y_test4.append(y_test3);
         
 
    button = Button(master, text="Process Dataset",height=1,fg="black",font=('algerian',13,'bold'),bg="violet",justify='center', command=lambda: process_dataset(var.get())) #Defining the button in the Tkinter Widget
    button.place(x=700,y=80)
 
    var1 = StringVar(master)
    var1.set("Select Classifier") # initial value
 
    option1 = OptionMenu(master, var1, "CNN", "Naive Bayes","K-Nearesr Neighbour")
    option1.place(x=500,y=120)
    option1.config(bg = "violet")
    option1.config(fg = "black")
    option1.config(font=('algerian',10,'bold'))
    option1.config(width=12)
    #option.place ( relx=0.5, rely=0.1)
    button1 = Button(master, text=" Train Classifier",height=1,fg="black",font=('algerian',13,'bold'),bg="violet",justify='center', command=lambda: train_classifier(x_train,x_test,y_train,y_test,var1.get()))
    button1.place(x=700,y=120)
 
 
 
 
    #e1.bind('<Button-1>',e1.delete(0,END))
     
    def predres(clf):                                                                           # Defining function to predict the result from the user input data   
        E14=E10.get()                                                                           # Converting the Eist data according to sign and decimal point
        if len(E14)==3 or len(E14)==4:
            E15=float(E14)
        else:
            E16=E14+'.0'
            E15=float(E16)
         
        test=[float(E1.get()+'.0'),float(E2.get()+'.0'),float(E3.get()+'.0'),float(E4.get()+'.0'),float(E5.get()+'.0'),float(E6.get()+'.0'),float(E7.get()+'.0'),float(E8.get()+'.0'),float(E9.get()+'.0'),E15,float(E11.get()),float(E12.get()+'.0'),float(E13.get()+'.0')]
        test=np.reshape(test,(1,-1))
        #print(test)
        print(clf)
        print(clf.predict(test))
        if clf.predict(test) < 0.49:
            res="The Person does not have heart Disease"
            Labx1=Label(master,text="The Person does not have heart Disease", bg='green')
            Labx1.visible=False
            Labx1.place(x=600,y=430)
            Labx1.visible=True
            #T3.insert(END,"The Person has Heart Disease")
 
        elif clf.predict(test) == 2:
            res="The Person has Severe heart Disease"
            Labx1=Label(master,text="The Person has Severe heart Disease", bg='red')
            Labx1.visible=False
            Labx1.place(x=600,y=430)
            Labx1.visible=True
         
        else:
            res="The Person has heart Disease"
            Labx1=Label(master,text="The Person has heart Disease           ", bg='red')
            Labx1.visible=False
            Labx1.place(x=600,y=430)
            Labx1.visible=True
        #T3.insert(END,"The Person does not have Heart Disease")
         
        age= E1.get()
        sex = E2.get()
        pai= E3.get()
        bp= E4.get()
        chol = E5.get()
        fbs = E6.get()
        ecg = E7.get()
        maxhr = E8.get()
        eiang = E9.get()
        eist = E10.get()
        slope = E11.get()
        vessels = E12.get()
        thal = E13.get()
        pana=E0.get()
 
        aa = mysql.connector.connect(host='localhost', port=3306, user="root", passwd="root", db="cardiac")
        mm = aa.cursor()
         
        mm.execute("""INSERT INTO cardiac1 VALUES (%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s)""", (age,sex,pai,bp,chol,fbs,ecg,maxhr,eiang,eist,slope,vessels,thal,pana,res))
        aa.commit()
        #con.close()
    def geterror(x_train,y_train,clf,title):
        #global clf   
        #global outclass
        import matplotlib.pyplot as plt
        from sklearn.model_selection import learning_curve
        from sklearn.model_selection import ShuffleSplit
        clas=[];
     
        def plot_learning_curve(estimator, title, X, y, ylim=None, cv=None,
                        n_jobs=1, train_sizes=np.linspace(.1, 1.0, 5)):
            plt.figure()
            plt.title(title)
            if ylim is not None:
                plt.ylim(*ylim)
            plt.xlabel("Training examples")
            plt.ylabel("Score")
            train_sizes, train_scores, test_scores = learning_curve(
                estimator, X, y, cv=cv, n_jobs=n_jobs, train_sizes=train_sizes)
            train_scores_mean = np.mean(train_scores, axis=1)
            train_scores_std = np.std(train_scores, axis=1)
            test_scores_mean = np.mean(test_scores, axis=1)
            test_scores_std = np.std(test_scores, axis=1)
            plt.grid()
 
            plt.fill_between(train_sizes, train_scores_mean - train_scores_std,
                train_scores_mean + train_scores_std, alpha=0.1,
                color="r")
            plt.fill_between(train_sizes, test_scores_mean - test_scores_std,
                test_scores_mean + test_scores_std, alpha=0.1, color="g")
            plt.plot(train_sizes, train_scores_mean, 'o-', color="r",
                 label="Training score")
            plt.plot(train_sizes, test_scores_mean, 'o-', color="g",
                 label="Cross-validation score")
 
            plt.legend(loc="best")
            plt.axis([ 0,len(y),0,1.1])
            return plt
 
     
        for index in range(len(y_train)):
            #x_train1=np.reshape(x_train.iloc[index:],(1,-1));
            x_train1=x_train.iloc[index].values.reshape(1,-1);
            outclass = clf.predict(x_train1);
            clas.append(outclass[0]);
         
        ind3 = np.where((clas == y_train));
        ind4 = np.where((clas != y_train));
        l=0
        m=0
     
        cv = ShuffleSplit(n_splits=4, test_size=0.2, random_state=0)
     
        plot_learning_curve(clf, title, x_train, y_train, (0.7, 1.01), cv=cv, n_jobs=1)
     
        plt.show()
         
     
     
    button2 = Button(master, text=" Predict Heart Disease ",width=20,height=1,fg="black",font=('algerian',13,'bold'),bg="violet",justify='center',command=lambda:predres(clf))
    button2.place(x=600,y=300)
    btn6=Button(master,text="LOGOUT",width=8,height=1,fg="black",font=('algerian',15,'bold'),bg="SKYBLUE",justify='center',command=cardes)
    btn6.place(x=1100,y=80)
    #button2.configure(width=14)
    #button1.place(relx=0.1,rely=0.2)
 
    master.mainloop()
 
 
def adminlogin():
    def adminlogininto():
        usernames = e1.get()
        passwords = e2.get()
        if e1.get() == "" or e2.get() == "":
            tkinter.messagebox.showinfo("sorry","Please complete the required field")
        elif e1.get() == "admin" and e2.get() == "admin":
            #tkMessageBox.showinfo("yeh","logged in")
            admindes()
        else:
            tkinter.messagebox.showinfo("Sorry" , "Wrong Password")
    global window1
    window1=Tk()
    window1.title("LOGIN PAGE")
     
    window1.geometry('700x500')
    image = Image.open('photo.png')
    image = image.resize((700, 600))
    photo_image = ImageTk.PhotoImage(image)
    label = Label(window1, image = photo_image)
    label.place(x=0,y=0)
     
    '''lb1=Label(window1,text="USERNAME",font=('algerian',25,'bold'),fg="BLACK",anchor='w')
    lb1.place(x=150,y=400)'''
 
    e1=Entry(window1,width=10,font=("bold",17),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    e1.place(x=250,y=150)
 
    '''lb2=Label(window1,text="PASSWORD",font=('algerian',25,'bold'),fg="BLACK",anchor='w')
    lb2.place(x=150,y=450)'''
 
    e2=Entry(window1,width=10,show="*",font=("bold",17),highlightthickness=2,bg="WHITE",relief=SUNKEN)
    e2.place(x=250,y=200)
 
    btn6=Button(window1,text="LOGIN",width=8,height=1,fg="black",font=('algerian',15,'bold'),bg="SKYBLUE",justify='center',command=adminlogininto)
    btn6.place(x=270,y=300)
 
    window1.mainloop()
 
def admindes():
    window1.destroy()
    cardiac()
 
def cardes():
    master.destroy()
    adminlogin()
     
if __name__ == "__main__":
    adminlogin()

            

        
      

Error:Exception in Tkinter callback
Traceback (most recent call last):
  File "C:\Users\Dell\AppData\Local\Programs\Python\Python37\lib\tkinter\__init__.py", line 1705, in __call__
    return self.func(*args)
  File "C:\Users\Dell\Desktop\finalproject\heartfinal.py", line 755, in <lambda>
    button2 = Button(master, text=" Predict Heart Disease ",width=20,height=1,fg="black",font=('algerian',13,'bold'),bg="violet",justify='center',command=lambda:predres(clf))
NameError: name 'clf' is not defined

jefsummers · Mar-31-2020, 08:14 AM

Global are a bad idea in general and this is part of why. Clf may be a global, but since you have a parameter called calf 100 lines higher a different clf is created. So it is a scope problem, as is common when working with global.

saiankith · Mar-31-2020, 08:27 AM

(Mar-31-2020, 08:14 AM)jefsummers Wrote: Global are a bad idea in general and this is part of why. Clf may be a global, but since you have a parameter called calf 100 lines higher a different clf is created. So it is a scope problem, as is common when working with global.

Thank you sir......so wat shld i do to solve dis

jefsummers · Mar-31-2020, 06:49 PM

Two options, either one will require you to go through your code in some detail.
1. Don't use clf as a global. That's what I would prefer. Pass it as a parameter when needed and return it when needed.
2. Use clf as a global and get rid of it in all function calls. If it is global it does not need to be passed.

To see some of the issues you can fall into with globals, look at the following code

        
              global clf
clf = 1
def fnc():
    clf = 3
    print(clf)
fnc()
print(clf)

Output:3
1

So how can clf be 3 inside the function but 1 outside?
Fix is to do this:

        
              global clf
clf = 1
def fnc():
    global clf
    clf = 3
    print(clf)
fnc()
print(clf)

Output:3
3

Now changing clf inside the function changes it outside the function.
That is SOOO prone to mistakes that are hard to find and code that is difficult to maintain. Which is why you need to go through your code and fix the globals one way or another.

Possibly Related Threads…
Thread		Author	Replies	Views	Last Post
	Variable is not defined error when trying to use my custom function code	fnafgamer239	4	1,602	Nov-23-2023, 02:53 PM Last Post: rob101
	[variable] is not defined error arises despite variable being defined	TheTypicalDoge	4	3,595	Apr-05-2022, 04:55 AM Last Post: deanhystad
	Error 'Contour' not Defined	DaveG	3	3,794	Mar-13-2022, 03:29 AM Last Post: deanhystad
	Getting "name 'get_weather' is not defined error and no json_data returned?	trthskr4	6	5,240	Sep-14-2021, 09:55 AM Last Post: trthskr4
	Error when refering to class defined in 'main' in an imported module	HeRo	2	3,253	Apr-13-2021, 07:22 PM Last Post: HeRo
	Why does lambda throw 'name value_o is not defined' error?	karabakh	3	3,198	Dec-14-2020, 05:45 PM Last Post: karabakh
	name error "name"is not defined	MaartenRo	1	5,162	Jul-28-2020, 02:39 AM Last Post: bowlofred
	Name Error: name 'Stockton' is not defined	Pinokchu	3	3,050	Jun-13-2020, 02:48 PM Last Post: Yoriz
	python library not defined in user defined function	johnEmScott	2	5,031	May-30-2020, 04:14 AM Last Post: DT2000
	error ,,name append is not defined''	Killdoz	1	6,113	May-24-2020, 06:23 PM Last Post: bowlofred

name error:name 'clf' is not defined

User Panel Messages

Announcements