How to use calc_label method in hypothesis

Best Python code snippet using hypothesis

script.py

Source:script.py

1'''2University at Buffalo - Spring 20153CSE 574 - Introduction to Machine Learning4Programming Assignment 25Authors: Dheeraj Balakavi, Pravin Umamaheswaran, Mithun Nagesh6It's not who you are, it's what you do that defines you - Batman / Bruce Wayne7'''8from __future__ import division9import numpy as np10from scipy.optimize import minimize11from scipy.io import loadmat12from math import sqrt13import scipy.io14import matplotlib.pyplot as plt15import pickle16def ldaLearn(X,y):17    # Inputs18    # X - a N x d matrix with each row corresponding to a training example19    # y - a N x 1 column vector indicating the labels for each training example20    #21    # Outputs22    # means - A d x k matrix containing learnt means for each of the k classes23    # covmat - A single d x d learnt covariance matrix 24    25    # IMPLEMENT THIS METHOD26    27    28    means = np.array([])29    unique_y = np.unique(y)30    31    i = 0;32    # processing each unique values in batches33    for index_val in unique_y:34        35        mapp_matrix = (y == index_val)36        mapp_matrix.shape = ((1,X.shape[0])) 37        #extended matrix38        ex_mapp_matrix = np.tile(mapp_matrix,X.shape[1])39        ex_mapp_matrix = np.reshape(ex_mapp_matrix,(X.shape[1],X.shape[0]))40        ex_mapp_matrix = ex_mapp_matrix.T41       42        MeanMat = np.asarray(X[ex_mapp_matrix])43        new_row = MeanMat.shape[0]/X.shape[1]44       45        KIK = np.reshape(MeanMat,(new_row,X.shape[1]))46        #taking the maen values of i/p per input data variable(x^i) per class47        Mean_Val = KIK.mean(0)48   49        if i == 0:50            means = np.hstack(Mean_Val)51        else :52            means = np.vstack((means,Mean_Val))  53        i+=1  54          55    means = means . T56    covmat = np.cov(X.T) # i am not sure about the covarience57        58    return means,covmat59def qdaLearn(X,y):60    # Inputs61    # X - a N x d matrix with each row corresponding to a training example62    # y - a N x 1 column vector indicating the labels for each training example63    #64    # Outputs65    # means - A d x k matrix containing learnt means for each of the k classes66    # covmats - A list of k d x d learnt covariance matrices for each of the k classes67    68    # IMPLEMENT THIS METHOD69    covmats = list()70    unique_classes = np.unique(y)71    mean_vectors = []72    for cl in range(1,5):73        mean_vectors.append(np.mean(X[y==cl], axis=0))74    75    i =0 76    for class_val in unique_classes:77         mapping_matrix = (y==class_val)78         mapping_matrix.shape = ((1,X.shape[0]))79         80         # creating an extended matrix by using tile function81         # to match the dimensions and separate by class82            83         ext_map_matrix = np.tile(mapping_matrix,X.shape[1])84         ext_map_matrix = np.reshape(ext_map_matrix,(X.shape[1],X.shape[0]))85         ext_map_matrix = ext_map_matrix.T86         87         class_ip = np.asarray(X[ext_map_matrix])88         row_size = class_ip.shape[0]/X.shape[1]89         90         class_ip = np.reshape(class_ip,(row_size,X.shape[1]))91         covmat = np.cov(class_ip.T)92         covmats.append(covmat)93       94         #take mean values of i/p per input data variable per class95         mean_value = class_ip.mean(0)96         if i == 0:97            means = np.hstack(mean_value)98         else :99            means = np.vstack((means,mean_value))  100         i+=1  101          102    means = means . T103  104    return means,covmats105def ldaTest(means,covmat,Xtest,ytest):106    # Inputs107    # means, covmat - parameters of the LDA model108    # Xtest - a N x d matrix with each row corresponding to a test example109    # ytest - a N x 1 column vector indicating the labels for each test example110    # Outputs111    # acc - A scalar accuracy value112    113    # IMPLEMENT THIS METHOD114    no_of_classes = 5115    inv_covariance = np.linalg.inv(covmat)  116    N = Xtest.shape[0]117   118    calc_label = np.array([])119    #calc_label = np.reshape(calc_label,(100,1))120    new_mean = means.T121    i = 0122    for each_row in range(N):   # for each row data calculate the probability or the classes123         row_data = Xtest[each_row,:]124         pdf_vector = np.array([])125         for each_class in range(no_of_classes):126             X_MU = row_data - new_mean[each_class]127             intmdt = np.dot(X_MU.T,inv_covariance)128             temp    = np.dot(intmdt,X_MU)129             d = 1130             pdf = np.exp(-1/2*temp)131             pdf_vector = np.append(pdf_vector,pdf)132             #pdf_vector[each_class] = pdf133         calc_label = np.append(calc_label,[np.argmax(pdf_vector)+1]) 134    calc_label  = np.matrix(calc_label).T      135   136    acc = 100 * np.mean((calc_label == ytest).astype(float)) 137    #plotGraph(Xtest,ytest,calc_label)138    return acc139def qdaTest(means,covmats,Xtest,ytest):140    # Inputs141    # means, covmats - parameters of the QDA model142    # Xtest - a N x d matrix with each row corresponding to a test example143    # ytest - a N x 1 column vector indicating the labels for each test example144    # Outputs145    # acc - A scalar accuracy value146    147    # IMPLEMENT THIS METHOD148    #no_of_classes = 5149    k = 5150    N = Xtest.shape[0]151    i= 0152    new_mean = means.T153    calc_label = np.array([])154    for each_row in range(N):  # for each row of the data calculate the probability 155      x_i = Xtest[each_row,:]156      pdf_vector = np.array([])157      for cl,cov_mat in zip(range(1,6),covmats):158          inv_covmat = np.linalg.inv(cov_mat) # inverse covariance matrix159          normlz_factor = 1/np.sqrt((2*np.pi)**k * np.linalg.det(cov_mat))160         161          X_MU   = x_i - new_mean[cl-1]162          intmdt = np.dot(X_MU.T,inv_covmat)163          temp   = np.dot(intmdt,X_MU)164          pdf    = np.exp(-1/2*temp)*normlz_factor165          pdf_vector = np.append(pdf_vector,pdf)166      calc_label = np.append(calc_label,[np.argmax(pdf_vector)+1])167    168    calc_label = np.matrix(calc_label).T      169    170    acc = 100 * np.mean((calc_label == ytest).astype(float)) 171    172    return acc173# Author: Dheeraj Balakavi174    # Inputs:                                                         175    # X = N x d 176    # y = N x 1                                                               177    # Output: 178    # w = d x 1179def learnOLERegression(X,y):180    # Step-1: Compute the Pseudo-Inverse of X --> PI(X)181    XT = X.transpose()182    XT_into_X = np.dot (XT, X)183    XT_into_X_inv = np.linalg.inv(XT_into_X)184    185    X_pi = np.dot (XT_into_X_inv, XT)186    187    # Test to see if I got it right.. if we do (X-pi * x) we must get an identity matrix 188    # as it is pseudo inverse we are doing189    X_ident = np.dot (X_pi, X)190    191    # Step-2: Compute w192    w = np.dot (X_pi, y)                           193    return w194# Author: Dheeraj Balakavi195    # Inputs:196    # X = N x d                                                               197    # y = N x 1 198    # lambd = ridge parameter (scalar)199    # Output:                                                                  200    # w = d x 1   201def learnRidgeRegression(X,y,lambd):202    X_rows, X_columns = X.shape203    N = X_rows204    M = X_columns205    # Step-1: Compute the Pseudo-Inverse of X with regularization co-eff206    XT = X.transpose()207    XT_into_X = np.dot (XT, X)208    identity = np.identity (M)209    lambd_into_NI = (lambd * N )* identity210    211    sum_lambNI_XTX = lambd_into_NI + XT_into_X212    sum_lambNI_XTX_inv = np.linalg.inv(sum_lambNI_XTX)213    214    # Step-2: Compute w215    product_one = np.dot (sum_lambNI_XTX_inv, XT)216    w = np.dot (product_one, y)217                                                218    return w219# Author: Dheeraj Balakavi220    # Inputs:221    # w = d x 1222    # Xtest = N x d223    # ytest = X x 1224    # Output:225    # rmse226def testOLERegression(w,Xtest,ytest):227    # Step-1: Compute the root mean square error (rmse)228    X_rows, X_columns = Xtest.shape229    N = X_rows230    231    wT = w.transpose()232    wT_into_Xtest = np.dot (Xtest, wT.transpose())233    234    subVal = np.subtract (ytest, wT_into_Xtest)235    subVal_sqr = subVal * subVal236    237    sumVal = np.sum (subVal_sqr)238    239    sumVal_sqrt = np.sqrt(sumVal)240    rmse = sumVal_sqrt / N241    return rmse242# Author: Dheeraj Balakavi243    # compute squared error (scalar) and gradient of squared error with respect244    # to w (vector) for the given data X and y and the regularization parameter245    # lambda 246def regressionObjVal(w, X, y, lambd):247    # Step-1: Error computation --> scalar248    X_rows, X_columns = X.shape249    N = X_rows250    w = np.vstack (w)251    #w = np.reshape(w, (w.shape[0], 1))    252    Xw = np.dot (X, w)253    y_minus_Xw = y - Xw254    y_minus_Xw_transpose = y_minus_Xw.transpose()255     256    prod1 = (np.dot (y_minus_Xw_transpose, y_minus_Xw)) / (2 * N)   257    prod2 = (1/2) * lambd * np.dot (w.transpose(), w)    258    error = (prod1 + prod2).item(0)259    #print error260    # Step-2: Error grad computation --> vector261    XT_X = np.dot (X.transpose(), X)262    wT_XT_X = np.dot (w.transpose(), XT_X)  263    yT_X = np.dot (y.transpose(), X)264    yT_X = -1 * yT_X265    266    sum1 = (yT_X + wT_XT_X) / N267    sum2 = lambd * w.transpose()268    error_grad = sum1 + sum2269    error_grad = np.squeeze(np.asarray(error_grad))270    #print error_grad                         271    272    return error, error_grad273def mapNonLinear(x,p):274    # Inputs:                                                                  275    # x - a single column vector (N x 1)                                       276    # p - integer (>= 0)                                                       277    # Outputs:                                                                 278    # Xd - (N x (d+1))                                                         279    # IMPLEMENT THIS METHOD280    output = []281    for val in np.nditer(x):282        283        #if p == 0:284        #    Xd = np.ones((x.shape[0],1))285        #    break286        287        for power  in range(0,p+1):288            #print 'awesome'            289            output.append(pow(val,power))290    #print 'x:' + str(x.shape[0])291    #print 'p:' + str(p)292    #print 'Size is:' + str(len(output))293    Xd = np.asarray(output)294    Xd = np.reshape(Xd,(x.shape[0],p+1))295    return Xd296# Main script297# Problem 1298# load the sample data 299                                                               300X,y,Xtest,ytest = pickle.load(open('sample.pickle','rb'))            301# LDA302means,covmat = ldaLearn(X,y)303ldaacc = ldaTest(means,covmat,Xtest,ytest)304print('LDA Accuracy = '+str(ldaacc))305# QDA306means,covmats = qdaLearn(X,y)307qdaacc = qdaTest(means,covmats,Xtest,ytest)308print('QDA Accuracy = '+str(qdaacc))309# Problem 2310X,y,Xtest,ytest = pickle.load(open('diabetes.pickle','rb'))   311# add intercept312X_i = np.concatenate((np.ones((X.shape[0],1)), X), axis=1)313Xtest_i = np.concatenate((np.ones((Xtest.shape[0],1)), Xtest), axis=1)314w = learnOLERegression(X,y)315mle = testOLERegression(w,Xtest,ytest)316w_i = learnOLERegression(X_i,y)317mle_i = testOLERegression(w_i,Xtest_i,ytest)318print('RMSE without intercept '+str(mle))319print('RMSE with intercept '+str(mle_i))320# Problem 3321k = 101322lambdas = np.linspace(0, 0.004, num=k)323i = 0324rmses3 = np.zeros((k,1))325for lambd in lambdas:326    w_l = learnRidgeRegression(X_i,y,lambd)327    rmses3[i] = testOLERegression(w_l,Xtest_i,ytest)328    i = i + 1329#plt.plot(lambdas,rmses3)330# Problem 4331k = 101332lambdas = np.linspace(0, 0.004, num=k)333i = 0334rmses4 = np.zeros((k,1))335opts = {'maxiter' : 100}    # Preferred value.                                                336w_init = np.zeros((X_i.shape[1],1))337for lambd in lambdas:338    args = (X_i, y, lambd)339    w_l = minimize(regressionObjVal, w_init, jac=True, args=args,method='CG', options=opts)340    w_l_1 = np.zeros((X_i.shape[1],1))341    for j in range(len(w_l.x)):342        w_l_1[j] = w_l.x[j]343    rmses4[i] = testOLERegression(w_l_1,Xtest_i,ytest)344    i = i + 1345#plt.plot(lambdas,rmses4)346fig, ax = plt.subplots()347fig.suptitle('Using Gradient Descent for Ridge Regression Learning', fontsize=14, fontweight='bold')348ax.set_title('Lambda vs RMSE4 for Training data')349ax.plot(lambdas,rmses4)350ax.set_xlabel('Lambda')351ax.set_ylabel('RMSE4')352ax.legend(loc='upper right')353# Problem 5354pmax = 7355lambda_opt = lambdas[np.argmin(rmses4)]356rmses5 = np.zeros((pmax,2))357for p in range(pmax):358    Xd = mapNonLinear(X[:,2],p)359    Xdtest = mapNonLinear(Xtest[:,2],p)360    w_d1 = learnRidgeRegression(Xd,y,0)361    rmses5[p,0] = testOLERegression(w_d1,Xdtest,ytest)362    w_d2 = learnRidgeRegression(Xd,y,lambda_opt)363    rmses5[p,1] = testOLERegression(w_d2,Xdtest,ytest)364plt.plot(range(pmax),rmses5)...

main.py

Source:main.py

1# Temperature converter2from tkinter import *3from tkinter import messagebox4root = Tk()5root.geometry('800x600')6root.config(bg='#00A676')7# Frames for celsius to fahrenheit and vice versa8c_f_frame = Frame(root, width=300, height=200, bg='#333')9c_f_frame.place(x=80, y=50)10f_c_frame = Frame(root, width=300, height=200, bg='#333')11f_c_frame.place(x=420, y=50)12# Headings for celsius to fahrenheit and vice versa13c_f_label = Label(c_f_frame, text='Celsius to Fahrenheit', bg='#333', fg='white')14c_f_label.place(x=10, y=10)15f_c_label = Label(f_c_frame, text='Fahrenheit to Celsius', bg='#333', fg='white')16f_c_label.place(x=10,y=10)17# Entry fields for celsius to fahrenheit and vice versa18c_f_entry = Entry(c_f_frame, state='readonly')19c_f_entry.place(x=70, y=100)20f_c_entry = Entry(f_c_frame, state='readonly')21f_c_entry.place(x=70, y=100)22# Functionality for activating entry fields23def c_f_activation():24    c_f_entry.config(state='normal')25    c_f_entry.focus()26    f_c_entry.config(state='normal')27    f_c_entry.delete(0, END)28    f_c_entry.config(state='readonly')29def f_c_activation():30    f_c_entry.config(state='normal')31    f_c_entry.focus()32    c_f_entry.config(state='normal')33    c_f_entry.delete(0, END)34    c_f_entry.config(state='readonly')35# Entry activation buttons36c_f_activate = Button(root, text='Activate - Celsius to fahrenheit', command=c_f_activation)37c_f_activate.place(x=115, y=270)38f_c_activate = Button(root, text='Activate - Fahrenheit to Celsius', command=f_c_activation)39f_c_activate.place(x=455, y=270)40# Temperature conversion function41def conversion():42    try:43        # Clears result label44        calc_label.config(text='')45        # Celsius to Fahrenheit46        if c_f_entry['state'] == 'normal':47            c_temp = float(c_f_entry.get())48            f_result = round((c_temp * (9/5)) + 32, 1)49            calc_label.config(text=f_result)50        # Fahrenheit to Celsius51        elif f_c_entry['state'] == 'normal':52            f_temp = float(f_c_entry.get())53            c_result = round((f_temp-32)*(5/9), 1)54            calc_label.config(text=c_result)55    # raise value error if invalid entry given56    except ValueError:57        messagebox.showerror(message='Invalid entry')58# Convert button and label for result59calc = Button(root, text='Calculate Conversion', command=conversion)60calc.place(x=200, y=350)61calc_label = Label(root, width=20, height=2, bg='#333', fg='white')62calc_label.place(x=420, y=350)63# functionality for clear button to clear fields64def delete():65    c_f_entry.config(state='normal')66    f_c_entry.config(state='normal')67    calc_label.config(text='')68    c_f_entry.delete(0, END)69    f_c_entry.delete(0, END)70    f_c_entry.config(state='readonly')71    c_f_entry.config(state='readonly')72# clear button73clear = Button(root, text='Clear', command=delete)74clear.place(x=550, y=500)75# Exit button with inbuilt function76escape = Button(root, text='Exit', command='exit')77escape.place(x=650, y=500)...

Automation Testing Tutorials

Learn to execute automation testing from scratch with LambdaTest Learning Hub. Right from setting up the prerequisites to run your first automation test, to following best practices and diving deeper into advanced test scenarios. LambdaTest Learning Hubs compile a list of step-by-step guides to help you be proficient with different test automation frameworks i.e. Selenium, Cypress, TestNG etc.