Numpy

Post 03 — Sorting methods

import numpy as npnp.set_printoptions(suppress=True)data = np.genfromtxt(r'./Numpy_Datasets/Lifecyclesavings.csv', delimiter=',', skip_header=1)
print(data.ndim)
print(data.shape)
print(data[:5,:])2
(50, 5)
[[  11.43   29.35    2.87 2329.68    2.87]
 [  12.07   23.32    4.41 1507.99    3.93]
 [  13.17   23.8     4.43 2108.47    3.82]
 [   5.75   41.89    1.67  189.13    0.22]
 [  12.88   42.19    0.83  728.47    4.56]]dt = {'names':["sr","pop15","pop75","dpi","ddpi"],
     'formats':[float, float, float, float, float]}
data.dtype=dt
print(data.ndim)
print(data.shape)2
(50, 1)data[:5]array([[(11.43, 29.35, 2.87, 2329.68, 2.87)],
       [(12.07, 23.32, 4.41, 1507.99, 3.93)],
       [(13.17, 23.8 , 4.43, 2108.47, 3.82)],
       [( 5.75, 41.89, 1.67,  189.13, 0.22)],
       [(12.88, 42.19, 0.83,  728.47, 4.56)]],
      dtype=[('sr', '<f8'), ('pop15', '<f8'), ('pop75', '<f8'), ('dpi', '<f8'), ('ddpi', '<f8')])# sort data inplace row-wise
data.sort(axis=0)
data[:5]array([[(0.6 , 39.74, 1.34,  662.86, 2.67)],
       [(1.27, 34.03, 3.08, 1900.1 , 1.12)],
       [(2.02, 41.18, 1.05,  220.56, 1.03)],
       [(2.81, 46.12, 1.21,  249.87, 1.13)],
       [(3.01, 46.05, 0.87,  289.71, 1.48)]],
      dtype=[('sr', '<f8'), ('pop15', '<f8'), ('pop75', '<f8'), ('dpi', '<f8'), ('ddpi', '<f8')])# descending sort
data.sort(axis=0)
data[::-1][:5]array([[(21.1 , 27.01, 1.91, 1257.28, 8.21)],
       [(18.56, 45.25, 0.56,  138.33, 5.14)],
       [(16.85, 24.42, 3.93, 2496.53, 3.99)],
       [(15.48, 32.54, 2.47,  601.05, 8.12)],
       [(14.65, 24.71, 3.25, 1740.7 , 7.66)]],
      dtype=[('sr', '<f8'), ('pop15', '<f8'), ('pop75', '<f8'), ('dpi', '<f8'), ('ddpi', '<f8')])#Lets reimport data
data = np.genfromtxt(r'./Numpy_Datasets/Lifecyclesavings.csv', delimiter=',', skip_header=1)
print(data.ndim)
print(data.shape)
print(data[:5,:])2
(50, 5)
[[  11.43   29.35    2.87 2329.68    2.87]
 [  12.07   23.32    4.41 1507.99    3.93]
 [  13.17   23.8     4.43 2108.47    3.82]
 [   5.75   41.89    1.67  189.13    0.22]
 [  12.88   42.19    0.83  728.47    4.56]]# first argsort on basis of first col to get index of elements , then using that index display first 5 rows of data
data[np.argsort(data[:,0])][:5]array([[   0.6 ,   39.74,    1.34,  662.86,    2.67],
       [   1.27,   34.03,    3.08, 1900.1 ,    1.12],
       [   2.02,   41.18,    1.05,  220.56,    1.03],
       [   2.81,   46.12,    1.21,  249.87,    1.13],
       [   3.01,   46.05,    0.87,  289.71,    1.48]])# reverse sort
np.argsort(data[:,0])array([ 6, 18, 31, 41, 16, 11, 23, 30, 49,  8, 40,  3, 38, 29, 43, 17, 46,
       42,  5, 48, 19, 44, 47, 26, 24, 28, 15,  9, 35, 12, 20,  0, 37,  7,
        1, 34, 14, 13, 32, 33,  4,  2, 36, 39, 21, 27, 25, 10, 45, 22],
      dtype=int64)data[np.argsort(data[:,0])[::-1]][:5]array([[  21.1 ,   27.01,    1.91, 1257.28,    8.21],
       [  18.56,   45.25,    0.56,  138.33,    5.14],
       [  16.85,   24.42,    3.93, 2496.53,    3.99],
       [  15.48,   32.54,    2.47,  601.05,    8.12],
       [  14.65,   24.71,    3.25, 1740.7 ,    7.66]])### lexsort sorting happens from right to left key columns 
data = data.round()
data[np.lexsort((data[:,1].tolist(), data[:,0].tolist()))][:5]array([[   1.,   34.,    3., 1900.,    1.],
       [   1.,   40.,    1.,  663.,    3.],
       [   2.,   41.,    1.,  221.,    1.],
       [   3.,   46.,    1.,  290.,    1.],
       [   3.,   46.,    1.,  250.,    1.]])# reimport 
data = np.genfromtxt(r'./Numpy_Datasets/Lifecyclesavings.csv', delimiter=',', skip_header=1)
data = data.round()

# sort 2nd index column ascending order and then 4th index column in descending order
data[np.lexsort([-data[:,4], data[:,2]])][:10]array([[ 12.,  45.,   1., 290.,   7.],
       [  4.,  42.,   1., 208.,   6.],
       [ 13.,  42.,   1., 728.,   5.],
       [ 19.,  45.,   1., 138.,   5.],
       [  5.,  47.,   1., 243.,   5.],
       [  4.,  44.,   1., 569.,   4.],
       [  1.,  40.,   1., 663.,   3.],
       [  5.,  47.,   1., 277.,   3.],
       [ 11.,  48.,   1., 471.,   3.],
       [  8.,  47.,   1., 232.,   3.]])