My compact solution…

import sys
import numpy as np

_ = input()  # We don't need N, M, or P.  Skip them!
print(np.loadtxt(sys.stdin, int))

This was a very, very poor problem. The challenge was to take the input, make two arrays with it, then concatenate them into one. However, if the input for the two arrays is instead read into a single array, there's no need to concatenate two arrays.

With that in mind, here's a more facetious solution that doesn't use NumPy, but passes all the tests…

import sys

t = map(str.strip, sys.stdin.readlines()[1:])
print('[[', ']\n ['.join(t), ']]', sep='')

😛

A better problem would've been to read in two arrays, then concatenate them in the opposite order.

import numpy as np

n,m,p = map(int,input().split()) ar1 = np.array([input().split() for i in range(n)], int) ar2 = np.array([input().split() for i in range(m)], int)

print(np.concatenate((ar1,ar2), axis=0))

Here is HackerRank Concatenate in Python solution - https://programmingoneonone.com/hackerrank-concatenate-problem-solution-in-python.html

I get what we're going for here, but I can't help but notice that none of the methods this was probably designed to test are actually necessary. There's no reason to concatenate, or handle the length of each array, nor worry about which axis it will be joined on, because the damn input is already formatted as a 2d array!

import numpy as np

# we don't need the first row, toss it:
_ = input()

data = []

# assign each input() result to a variable and append it to our empty list:
while True:
    try:
        line = input()
        data.append(list(map(int, line.split())))
    except EOFError:
        break

# change the list to an array and print it:
arr = np.array(data)
print(arr)

`