Created
July 11, 2021 11:10
-
-
Save devarshi16/5fd9266cf1fa83143088b1f630b363d3 to your computer and use it in GitHub Desktop.
Compare execution times of two different one hot encoding algorithms using numpy and python
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| import numpy as np | |
| import matplotlib.pyplot as plt | |
| import random | |
| import time | |
| def one_hot(Y): | |
| data_size=Y.shape[0] | |
| classes=np.unique(Y).reshape(-1,1) | |
| num_classes=classes.shape[0] | |
| class_mappings=np.arange(0,max(Y)+1) | |
| class_mappings[np.unique(classes)]=np.arange(num_classes) | |
| Y=class_mappings[Y] | |
| one_hot=np.zeros((data_size,num_classes)) | |
| #rows=np.arange(data_size) | |
| one_hot[np.arange(data_size).reshape(-1,1),Y.reshape(-1,1)]=1 | |
| class_col=np.sort(classes) | |
| return one_hot,class_col | |
| def one_hot_for(Y): | |
| data_size=Y.shape[0] | |
| classes=np.unique(Y).reshape(-1,1) | |
| num_classes=classes.shape[0] | |
| one_hot=np.zeros((data_size,num_classes)) | |
| for row in range(data_size): | |
| one_hot[row,np.where(classes==Y[row])[0]]=1 | |
| return one_hot,classes | |
| # Generate a randoms file | |
| file_name="randoms.txt" | |
| with open(file_name,"w+") as random_labels: | |
| for i in range(10000): | |
| random_labels.write(str(random.randint(0,1000))+"\n") | |
| with open(file_name,"r+") as f: | |
| Y=f.readlines() | |
| int_map=map(int,Y) | |
| Y=list(int_map) | |
| Y=np.asarray(Y).reshape(-1,1) | |
| one_hot_timings=[] | |
| one_hot_for_timings=[] | |
| for i in range(100,10000,100): | |
| start=time.time() | |
| _,_=one_hot(Y[:i]) | |
| end=time.time() | |
| one_hot_timings.append(end-start) | |
| start=time.time() | |
| _,_=one_hot_for(Y[:i]) | |
| end=time.time() | |
| one_hot_for_timings.append(end-start) | |
| plt.plot(one_hot_timings,label="one_hot_vector") | |
| plt.plot(one_hot_for_timings,label="one_hot_for") | |
| plt.xlabel('data_size for every 100 datapoints') | |
| plt.ylabel('time of execution') | |
| plt.legend(loc='best') | |
| plt.show() |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment