So I have an assignment to code Stoachastic gradient decent and basically i am finding it a bit of a problem to randomly sample from multiple vectors while keeping the order intact. My code follows:
import numpy as np
import matplotlib.pyplot as plt
import random
x = np.array([0.,0.,0.,100.,100.,300.,300.,900.,900.,900.])
y = np.array([0.,0.,1.,0.,1.,1.,1.,0.,1.,1.])
def f(b0,b1,x,y):
vec = [y[i]*np.log(1/(1+np.exp(-b0-b1*x[i]))) + (1-y[i])*np.log(1 - (1/(1+np.exp(-b0-b1*x[i])))) for i in range(len(y))]
return sum(vec)
def dervf0(b0,b1,x,y):
vec = [-y[i] + (1/(1+np.exp(-b0-b1*x[i]))) for i in range(len(y))]
return sum(vec)
def dervf1(b0,b1,x,y):
vec = [-x[i]*(y[i]-(1/(1+np.exp(-b0-b1*x[i])))) for i in range(len(y))]
return sum(vec)
def SGD(v,x,y,tol,maxiter):
x = #random selection
y= #random selection
for i in range(maxiter):
theta_new = v - 0.001*np.array(
[dervf0(v[0], v[1], x, y),
dervf1(v[0], v[1], x, y)])
if np.linalg.norm(theta_new - v) < tol:
break
else:
v = theta_new
#print('i\t{}\tv\t{}\ttheta_new\t{}'.format(i, v, theta_new))
return theta_new,i
As you can see I have 2 vectors, x and y, and they are linked for example x[0] is an experiment which gave us y[0] = 0. It makes no sense to randomly sample without structure when here in my opinion. What i am struggling to do is in the SGD function, where i want n-points of x and n-points of y but structured correctly! any help is appreciated!
Y
You can get a list of indices to sample using the following-
import random
x = ['This', 'is', 'a', 'random', 'sampling', 'example']
n = len(x)
k = 5
indices_to_sample = sorted(random.sample(range(n),k)) # Chooses k out of n indices and sorts them
for i in indices_to_sample:
print(x[i]) # Gets x at index i
Read more at random.sample docs