Create 3D array from existing 2D array in Python

Question

I am trying to create an array that is with dimensions:

a(Days,Hours,Station)

I have hourly data for array 'a' for 2 stations over 61 days so currently I have this array with these dimensions:

a(1464,2)

Where the 1464 is the number of hourly data points per station that I have (24 hours*61 days). However I want to break it down even further and add another dimension that has Days so the dimensions would then be:

a(61 days,24 hours/day, 2 stations)

Any ideas on how I would correctly be able to take the array 'a' that I currently have and change it into these 3 dimensions?

a is numpy array or standard array?

Arnial
– Arnial

2016-07-05 21:14:16 +00:00
Commented Jul 5, 2016 at 21:14 — Arnial
– Arnial, Commented Jul 5, 2016 at 21:14
a is a numpy array.

HM14
– HM14

2016-07-05 21:51:13 +00:00
Commented Jul 5, 2016 at 21:51 — HM14
– HM14, Commented Jul 5, 2016 at 21:51

Arnial · Accepted Answer · 2016-07-05 20:23:42Z

1

This will split array a to chunks with maximum length size.

def chunks( a, size ):
    arr = iter( a )
    for v in arr:
        tmp = [ v ]
        for i,v in zip( range( size - 1 ), arr ):
            tmp.append( v )

        yield tmp

splitted = list( chunks( a, 24 ) )

answered Jul 5, 2016 at 20:23

Arnial

1,4611 gold badge11 silver badges10 bronze badges

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

5 Comments

HM14 Over a year ago

Oh ok interesting. I suppose I could do this then. I also read about np.rearrage(a, (61,2,2)) so I tried that but I don't exactly know what the function is doing behind the scenes. Are you familiar with this? Would this be an easy solution too?

HM14 Over a year ago

Also this doesn't quite work. Splitted is of dimensions (1,2,1464) instead of (61,24,2) like I want.

Arnial Over a year ago

Can't find any info on np.rearrage. Don't know anything abount it.

Arnial Over a year ago

Is it possible that shape of a is (2, 1464)? It's only possible reason, that i can find, that would lead to (1, 2, 1646).?

HM14 Over a year ago

Yes that is why! So I need to switch the dimensions around somehow and then if I do that and use this code it works. Thanks! Meghan

kcborys · Accepted Answer · 2016-07-05 19:46:35Z

1

You could try to make a 61x24x2 array. This should work:

b = []
for i in xrange(61):
    b.append(a[i*61:(i+1)*61])

answered Jul 5, 2016 at 19:46

kcborys

3161 silver badge11 bronze badges

Comments

Zach King · Accepted Answer · 2016-07-05 19:59:13Z

0

If you're first field is hours * days, a transformation would simply be: a(x, y) => a(x // 24, x % 24, y)

x // 24 is floor division so 1500 // 24 = 62, the days. You did not specify, but I assume the "Hours" field would be the remaining hours; so x % 24 gets the remaining hours, and 1500 % 25 = 12, the number of hours. Lastly, the station field remains the same.

I don't think you can modify the structure of a list/array in Python, so you would need to create a new one. I'm also not sure if you're actually using the built-in list or the array. I'm not too familiar with the array class so this isn't a complete answer, but I hope it points you in the right direction arithmetically.

answered Jul 5, 2016 at 19:59

Zach King

1,2781 gold badge11 silver badges21 bronze badges

Collectives™ on Stack Overflow

Create 3D array from existing 2D array in Python

3 Answers 3

5 Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

5 Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related