python - Find occurrences of a value in a numpy array and assign it appropriate weights -

i have text file of close 1 million lines.it has 2 columns.column 1 has numbers 0-99 , column has 4 sizes ranging s,m,l,xl. numbers 0 99 keep repeating in 1 million lines different sizes follows:

11 s 19 s 19 m 19 m 63 l 14 s 11 l 63 xl 14 s 11 l 63 xl 

my objective find final size each number.the plan of action find occurrence of each number,find size each occurrence , assign final size number maximum occurrences of size.

expected output :

11 l 14 s 19 m 63 xl 

because of size of dataset,i having @ numpy,not have prior experience it. have started creating basic numpy array follows:

import numpy np  data = np.loadtxt('size_data.txt') 

this create numpy array.however,from whatever documentation have read till now,it not there direct way of doing want accomplish.can 1 give me pointers on how move forward ?

we can applying numpy.unique on reversed version of first column received file. reversing required because other wise it(return_index=true) return indices of first occurrence found of item start.

>>> arr = np.loadtxt('foo.txt', dtype=object) >>> _, indices = np.unique(arr[:, 0][::-1], return_index=true) >>> arr[::-1][indices] array([['11', 'l'],        ['14', 's'],        ['19', 'm'],        ['63', 'xl']], dtype=object) # or >>> arr[len(arr) - indices - 1] array([['11', 'l'],        ['14', 's'],        ['19', 'm'],        ['63', 'xl']], dtype=object)