Skip to content

Instantly share code, notes, and snippets.

@sunilkumardash9
Last active January 30, 2022 11:43
Show Gist options
  • Save sunilkumardash9/21f95931882e16c268048ab4f72ef23c to your computer and use it in GitHub Desktop.
Save sunilkumardash9/21f95931882e16c268048ab4f72ef23c to your computer and use it in GitHub Desktop.
import numpy as np
from numpy import log,dot,exp,shape
import matplotlib.pyplot as plt
from sklearn.datasets import make_classification
X,y = make_classification(n_featues=4)
from sklearn.model_selection import train_test_split
X_tr,X_te,y_tr,y_te = train_test_split(X,y,test_size=0.1
def standardize(X_tr):
for i in range(shape(X_tr)[1]):
X_tr[:,i] = (X_tr[:,i] - np.mean(X_tr[:,i]))/np.std(X_tr[:,i])
def F1_score(y,y_hat):
tp,tn,fp,fn = 0,0,0,0
for i in range(len(y)):
if y[i] == 1 and y_hat[i] == 1:
tp += 1
elif y[i] == 1 and y_hat[i] == 0:
fn += 1
elif y[i] == 0 and y_hat[i] == 1:
fp += 1
elif y[i] == 0 and y_hat[i] == 0:
tn += 1
precision = tp/(tp+fp)
recall = tp/(tp+fn)
f1_score = 2*precision*recall/(precision+recall)
return f1_score
class LogidticRegression:
def sigmoid(self,z):
sig = 1/(1+exp(-z))
return sig
def initialize(self,X):
weights = np.zeros((shape(X)[1]+1,1))
X = np.c_[np.ones((shape(X)[0],1)),X]
return weights,X
def fit(self,X,y,alpha=0.001,iter=400):
weights,X = self.initialize(X)
def cost(theta):
z = dot(X,theta)
cost0 = y.T.dot(log(self.sigmoid(z)))
cost1 = (1-y).T.dot(log(1-self.sigmoid(z)))
cost = -((cost1 + cost0))/len(y)
return cost
cost_list = np.zeros(iter,)
for i in range(iter):
weights = weights - alpha*dot(X.T,self.sigmoid(dot(X,weights))-np.reshape(y,(len(y),1)))
cost_list[i] = cost(weights)
self.weights = weights
return cost_list
def predict(self,X):
z = dot(self.initialize(X)[1],self.weights)
lis = []
for i in self.sigmoid(z):
if i>0.5:
lis.append(1)
else:
lis.append(0)
return lis
standardize(X_tr)
standardize(X_te)
obj1 = LogidticRegression()
model= obj1.fit(X_tr,y_tr)
y_pred = obj1.predict(X_te)
y_train = obj1.predict(X_tr)
#Let's see the f1-score for training and testing data
f1_score_tr = F1_score(y_tr,y_train)
f1_score_te = F1_score(y_te,y_pred)
print(f1_score_tr)
print(f1_score_te)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment