How to use test_sequence method in molecule

Best Python code snippet using molecule_python

fetch_HMM.py

Source:fetch_HMM.py

1# Fetch the data from Firebase2from firebase import firebase3import math4firebase = firebase.FirebaseApplication('https://androidbletutorial.firebaseio.com/',None)5All_Sample = []6result=firebase.get('',None)7length = 108# Define Partition Function9def Partition(value):10    end_value = pow(2,16)-1  # maximam value11    global length12    num_level = length13    return_value = math.ceil(value*num_level/end_value)14    #print(value,"  ", return_value);15    return return_value16# Sample value17for key in result.keys():18  temp = result[key]["text"].split(":")[7]19  x1, x2 = temp.split(" ")[1], temp.split(" ")[2]20  y1, y2 = temp.split(" ")[3], temp.split(" ")[4]21  z1, z2 = temp.split(" ")[5], temp.split(" ")[6]22  x = int(x1) + int(x2)<<823  y = int(y1) + int(y2)<<824  z = int(z1) + int(z2)<<825  sample = math.sqrt(x*x+y*y+z*z)26  sample = Partition(int(sample))27  All_Sample.append(sample)28Sample_list = ''.join([str(elem) for elem in All_Sample]) 29print(Sample_list)30# find transition matrix and emission matrix31#defining states and sequence symbols32import numpy as np33import decimal34from decimal import Decimal35import math36import time37#transition matrix38transition = np.array([[0.9,0.1],[0.1,0.9]])39#emission matrix40emission = np.array([[0.05,0.05,0.05,0.05,0.1,0.1,0.1,0.1,0.2,0.2],41                     [0.2,0.2,0.1,0.1,0.1,0.1,0.05,0.05,0.05,0.05]])42states = ['H','C']43states_dic = {'H':0, 'C':1}44#sequence_syms = {'1':0,'6600':1,'14000':2,'28000':3,'35000':4,'41000':5,'47000':6,'55000':7,'57000':8,'65000':9}45sequence_syms = {'0':0,'1':1,'2':2,'3':3,'4':4,'5':5,'6':6,'7':7,'8':8,'9':9}46sequence = ['0','1','2','3','4','5','6','7','8','9']47#sequence = ['1', '6600', '14000', '28000', '35000', '41000', '47000', '55000', '57000', '65000']48#test sequence49test_sequence = Sample_list50#test_sequence = ['47000','35000','65000','65000','41000','55000','57000','65000','47000','55000']51test_sequence = test_sequence52#probabilities of going to end state53end_probs = [0.1, 0.1]54#probabilities of going from start state55start_probs = [0.5, 0.5]56#function to find forward probabilities57def forward_probs():58    # node values stored during forward algorithm59    node_values_fwd = np.zeros((len(states), len(test_sequence)))60    for i, sequence_val in enumerate(test_sequence):61        for j in range(len(states)):62            # if first sequence value then do this63            if (i == 0):64                node_values_fwd[j, i] = start_probs[j] * emission[j, sequence_syms[sequence_val]]65            # else perform this66            else:67                values = [node_values_fwd[k, i - 1] * emission[j, sequence_syms[sequence_val]] * transition[k, j] for k in68                          range(len(states))]69                node_values_fwd[j, i] = sum(values)70    #end state value71    end_state = np.multiply(node_values_fwd[:, -1], end_probs)72    end_state_val = sum(end_state)73    return node_values_fwd, end_state_val74#function to find backward probabilities75def backward_probs():76    # node values stored during forward algorithm77    node_values_bwd = np.zeros((len(states), len(test_sequence)))78    #for i, sequence_val in enumerate(test_sequence):79    for i in range(1,len(test_sequence)+1):80        for j in range(len(states)):81            # if first sequence value then do this82            if (-i == -1):83                node_values_bwd[j, -i] = end_probs[j]84            # else perform this85            else:86                values = [node_values_bwd[k, -i+1] * emission[k, sequence_syms[test_sequence[-i+1]]] * transition[j, k] for k in range(len(states))]87                node_values_bwd[j, -i] = sum(values)88    #start state value89    start_state = [node_values_bwd[m,0] * emission[m, sequence_syms[test_sequence[0]]] for m in range(len(states))]90    start_state = np.multiply(start_state, start_probs)91    start_state_val = sum(start_state)92    return node_values_bwd, start_state_val93#function to find si probabilities94def si_probs(forward, backward, forward_val):95    si_probabilities = np.zeros((len(states), len(test_sequence)-1, len(states)))96    for i in range(len(test_sequence)-1):97        for j in range(len(states)):98            for k in range(len(states)):99                si_probabilities[j,i,k] = ( forward[j,i] * backward[k,i+1] * transition[j,k] * emission[k,sequence_syms[test_sequence[i+1]]] ) \100                                                    / forward_val101    return si_probabilities102#function to find gamma probabilities103def gamma_probs(forward, backward, forward_val):104    gamma_probabilities = np.zeros((len(states), len(test_sequence)))105    for i in range(len(test_sequence)):106        for j in range(len(states)):107            #gamma_probabilities[j,i] = ( forward[j,i] * backward[j,i] * emission[j,sequence_syms[test_sequence[i]]] ) / forward_val108            gamma_probabilities[j, i] = (forward[j, i] * backward[j, i]) / forward_val109    return gamma_probabilities110#performing iterations until convergence111for iteration in range(2000):112    print('\nIteration No: ', iteration + 1)113    # print('\nTransition:\n ', transition)114    # print('\nEmission: \n', emission)115    #Calling probability functions to calculate all probabilities116    fwd_probs, fwd_val = forward_probs()117    bwd_probs, bwd_val = backward_probs()118    si_probabilities = si_probs(fwd_probs, bwd_probs, fwd_val)119    gamma_probabilities = gamma_probs(fwd_probs, bwd_probs, fwd_val)120    # print('Forward Probs:')121    # print(np.matrix(fwd_probs))122    123    # print('Backward Probs:')124    # print(np.matrix(bwd_probs))125    #126    # print('Si Probs:')127    # print(si_probabilities)128    # print('Gamma Probs:')129    # print(np.matrix(gamma_probabilities))130    #caclculating 'a' and 'b' matrices131    a = np.zeros((len(states), len(states)))132    b = np.zeros((len(states), len(sequence_syms)))133    #'a' matrix134    for j in range(len(states)):135        for i in range(len(states)):136            for t in range(len(test_sequence)-1):137                a[j,i] = a[j,i] + si_probabilities[j,t,i]138            denomenator_a = [si_probabilities[j, t_x, i_x] for t_x in range(len(test_sequence) - 1) for i_x in range(len(states))]139            denomenator_a = sum(denomenator_a)140            if (denomenator_a == 0):141                a[j,i] = 0142            else:143                a[j,i] = a[j,i]/denomenator_a144    #'b' matrix145    for j in range(len(states)): #states146        for i in range(len(sequence)): #seq147            indices = [idx for idx, val in enumerate(test_sequence) if val == sequence[i]]148            numerator_b = sum( gamma_probabilities[j,indices] )149            denomenator_b = sum( gamma_probabilities[j,:] )150            if (denomenator_b == 0):151                b[j,i] = 0152            else:153                b[j, i] = numerator_b / denomenator_b154    print('\nMatrix a:\n')155    print(np.matrix(a.round(decimals=4)))156    print('\nMatrix b:\n')157    print(np.matrix(b.round(decimals=4)))158    transition = a159    emission = b160    new_fwd_temp, new_fwd_temp_val = forward_probs()161    print('New forward probability: ', new_fwd_temp_val)162    diff =  np.abs(fwd_val - new_fwd_temp_val)163    print('Difference in forward probability: ', diff)164    if (diff < 0.000000001):...

classify_nn.py

Source:classify_nn.py

1import operator2import math3import numpy as np4def distance(A, B):5	sum_ = 06	for i in range(len(A)):7		sum_ += (A[i]-B[i])**28	return sum_9def find_L2(sequence, test_sequence):10	assert(len(sequence.frames) == len(test_sequence.frames))11	dist = 012	for i in range(len(sequence.frames)):13		sframe = sequence.frames[i].frame14		tframe = test_sequence.frames[i].frame15		dist += distance(sequence.frames[i].frame, test_sequence.frames[i].frame)16	return dist17def classify_nn(test_sequence, training_gesture_sets):18	"""19	Classify test_sequence using nearest neighbors20	:param test_gesture: Sequence to classify21	:param training_gesture_sets: training set of labeled gestures22	:return: a classification label (an integer between 0 and 8)23	"""24	min_dist = float('inf')25	result = None26	for gs in training_gesture_sets:27		total = 028		for seq in gs.sequences:29			l2 = find_L2(seq, test_sequence)30			total += l231		avg = total/len(gs.sequences)32		if avg < min_dist:33			min_dist = avg34			result = gs.label...

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.