You apply a convolution module to the data: y=convolve(x,kernel,bias), where x is the input sequence, kernel is the convolution kernel, bias are the shift parameters for each output channel. Write a function calculate_conv_x_grad that finds the value of the derivative of the result given the input diff(y)/diff(x). import sys import ast import numpy as np def parse_array(s): return np.array(ast.literal_eval(s)) def read_array(): return parse_array(sys.stdin.readline()) def write_array(arr): print(repr(arr. tolist())) def calculate_conv_x_grad(x, y, kernel, bias): """ x - InLen x InChannels y - OutLen x OutChannels kernel - OutChannels x InChannels x KernelSize bias - OutChannels returns InLen x InChannels """ x = read_array() y = read_array() kernel = read_array() bias = read_array() result = calculate_conv_x_grad(x, y, kernel, bias) write_array(result) Use the following values for testing and debugging: sample input: [[0.5031766517322117, 0.30744410216949514], [0.04690208449415345, 0.322727131626243], [0.1388690574185909, 0.48576543724022325 ], [0.5260018011862109, 0.5859221562109312] 0.8974007607375208, 0.5713329992292489], [0.378989716528242, 0.49787928388753266]] [[1.5157583762374225, 0.9460413662192456, 0.9802340338281511], [1.5728362445918327, 0.996409724139607, 1.2530013664472253], [1 .9068174476481374, 1.430592927945995, 1.6704630594015581], [2.189768979209843, 2.3149543871163503, 2.1601629609824995], [2.8353 457102707083, 1.7422359297539565, 1.816707087141475], [2.0532913525958474, 1.9924093441385802, 2.3069493556139014]] [[[0.8077620147648772, 0.006392942850116379, 0.6080212915877307], [0.6288229869798402, 0.6410664904844843, 0.75419330562945] ], [[0.5355186530459589, 0.9211024178840701, 0.27725553497982014], [0.4507098181629161, 0.081570594016668, 0.8234980185346139]], [[0.0325944131753374, 0.7744753133142763, 0.05946983249285043], [0.7059580971549311, 0.7969953841197822 , 0.5257810951530107]]] [0.2579976950685653, 0.029957050945287222, 0.18958928880952108] sample output: [[1.3758750809861735, 1.7854909022976875], [3.0778457550346365, 3.305123370918622], [4.022592414095037, 5.4085957902356965], [4. 022592414095037, 5.4085957902356965], [4.022592414095037, 5.4085957902356965], [4.022592414095037, 5.4085957902356965], [2.646717 333108864, 3.623104887938009], [0.9447466590604012, 2.1034724193170744]]

import sys
import ast
import numpy as np


def parse_array(s):
     return np.array(ast.literal_eval(s))

def read_array():
     return parse_array(sys.stdin.readline())

def write_array(arr):
     print(repr(arr. tolist()))


def calculate_conv_x_grad(x, y, kernel, bias):
     """
     x - InLen x InChannels
     y - OutLen x OutChannels
     kernel - OutChannels x InChannels x KernelSize
     bias - OutChannels

     returns InLen x InChannels
     """

x = read_array()
y = read_array()
kernel = read_array()
bias = read_array()

result = calculate_conv_x_grad(x, y, kernel, bias)

write_array(result)