Adding an extra in column into 2D numpy array python

Question

I have a 2D numpy array that has a shape of (867, 43). My aim is to add an extra column (np.nan value) as the leading column to this array so that the shape becomes (867, 44).

An example would be:

# sub-section of array
>>> arr[:2, :5]

array([[-0.30368954,  2.8808107 ,  5.8833385 ,  8.6606045 , 11.242557  ],
       [-0.22719575,  3.0030012 ,  6.065371  ,  8.924864  , 11.561942  ]],
  dtype=float32)

would turn into:

# same sub-section
>>> f[:2,:5]

array([[        nan, -0.30368954,  2.8808107 ,  5.8833385 ,  8.6606045 ],
       [        nan, -0.22719575,  3.0030012 ,  6.065371  ,  8.924864  ]],
  dtype=float32)

i.e the values have been shifted right as the horizontal dimension has increased by one.

rahlf23 · Accepted Answer · 2018-12-13 20:14:54Z

1

You can use np.hstack():

import numpy as np

my_arr = np.array([[-0.30368954,  2.8808107 ,  5.8833385 ,  8.6606045 , 11.242557  ],
                [-0.22719575,  3.0030012 ,  6.065371  ,  8.924864  , 11.561942  ]])

col = np.empty((my_arr.shape[0],1))
col[:] = np.nan
np.hstack((col, my_arr))

Returns:

[[        nan -0.30368954  2.8808107   5.8833385   8.6606045  11.242557  ]
 [        nan -0.22719575  3.0030012   6.065371    8.924864   11.561942  ]]

answered Dec 13, 2018 at 20:14

rahlf23

9,0394 gold badges30 silver badges57 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

cvanelteren · Accepted Answer · 2018-12-13 20:22:21Z

1

Have a look at stack. Edit: clarification; I am making use of the broadcasting feature to insert a newaxis along the second dimension and hstack will then append the axis along the zero axis (default for hstack is rows or first dimension).

from numpy import array, hstack, nan, newaxis
a = array([[-0.30368954,  2.8808107 ,  5.8833385 ,  8.6606045 , 11.242557  ],
       [-0.22719575,  3.0030012 ,  6.065371  ,  8.924864  , 11.561942  ]],
  dtype=float32)

tmp = ones((a.shape[0])) * nan # create nan array
print(hstack((tmp[:, newaxis], a))) # append along zero axis

Output:

[[        nan -0.30368954  2.88081074  5.88333845  8.66060448 11.24255657]
 [        nan -0.22719575  3.00300121  6.06537104  8.92486382 11.5619421 ]]

edited Dec 13, 2018 at 20:22

answered Dec 13, 2018 at 20:12

cvanelteren

1,70311 silver badges18 bronze badges

2 Comments

Max Collier Over a year ago

that works! However I'm sure there would be a way to do it which is a bitter clearer to whats going on... I'm trying to see if I can get numpy.insert to work in one line but I haven't got it to work yet..

cvanelteren Over a year ago

addded some more info for you! And cleaned it up with the required nan array. Either using the newaxis or None. It is equivalent to specify a 1 in the shape of the ones method.

Max Collier · Accepted Answer · 2018-12-13 20:25:33Z

0

Using np.insert()

>>> import numpy as np
>>> arr
array([[-0.30368954,  2.8808107 ,  5.8833385 ,  8.6606045 , 11.242557  ],
       [-0.22719575,  3.0030012 ,  6.065371  ,  8.924864  , 11.561942  ]],

>>> arr = np.insert(arr, 0, np.nan, axis=0)
>>> arr
array([[        nan, -0.30368954,  2.8808107 ,  5.8833385 ,  8.6606045 , 11.242557  ],
       [        nan, -0.22719575,  3.0030012 ,  6.065371  ,  8.924864  , 11.561942  ]],

answered Dec 13, 2018 at 20:25

Max Collier

6919 silver badges27 bronze badges

2 Comments

cvanelteren Over a year ago

Nice! Didn't know about insert. Normally use stack for all this stuff. Learning something new everyday

Max Collier Over a year ago

No problem! Your stack example was going to be my fallback option. Cheers for the help!

Collectives™ on Stack Overflow

Adding an extra in column into 2D numpy array python

3 Answers 3

Comments

2 Comments

2 Comments

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

2 Comments

2 Comments

Related