Numpy array of alternating tuples

Question

I am looking for a pythonic/simple way to create a list/sequence of repeating tuples (specifically RGBA values) so that I can apply it to the colors of bars in a matplotlib bar chart.

ax.bar(...
       , color=colorlist
   )

For example

print(colorlist)
array([(0.04, 0.015, 0.02, 1.0),
       (0.8, 0.5, 0.2, 1.0),
       (0.04, 0.015, 0.02, 1.0), ...,
       (0.8, 0.5, 0.2, 1.0),
       (0.04, 0.015, 0.02, 1.0),
       (0.8, 0.5, 0.2, 1.0)], dtype=object)

I know the length of colorlist (e.g. N) I am after. I just can't seem to find a way to create colorlist.

I am after a way to arrange the two colors (say color1 and color2) either one immediately followed by the second color (e.g. 0,1,0,1,0,1,..), or some number of one color in a row, then followed another number of the second color, and repeated until the end (0,0,0,1,1,1,0,0,0,1,1,1,..).

EDIT: A worked example;

import numpy as np
import matplotlib.pyplot as plt

colorlist = np.array(
   ([(0.04, 0.015, 0.02, 1.0)] * 1 + [(0.8, 0.5, 0.2, 1.0)] * 1) * 16, dtype=object
)

fig, ax = plt.subplots()
y = np.ones(32)
x = np.linspace(1,32,32)
ax.bar(x, y, facecolor = colorlist)

with error

ValueError: Invalid RGBA argument: array([[0.04, 0.015, 0.02, 1.0],
       [0.8, 0.5, 0.2, 1.0],
       [0.04, 0.015, 0.02, 1.0],```

Grismar · Accepted Answer · 2021-09-14 12:22:21Z

1

Something like:

colorlist = [(0.04, 0.015, 0.02, 1.0), (0.8, 0.5, 0.2, 1.0)] * 10

Gets you 10 repetitions of the list as a list, which you can then cast to an array, as needed?

Similarly, if you want to repeat each colour a known number of times:

colorlist = ([(0.04, 0.015, 0.02, 1.0)] * 2 + [(0.8, 0.5, 0.2, 1.0)] * 3) * 4

Gets you 4 repetitions of the first colour twice and the second colour three times.

To cast this to an array of object, for ax.bar, you can do something like:

import numpy as np


colorlist = np.array(
    ([(0.04, 0.015, 0.02, 1.0)] * 2 + [(0.8, 0.5, 0.2, 1.0)] * 3) * 4, dtype=object
)

Given the specific example code, using color instead of facecolor works with the original list, even without the cast or the array data type:

import matplotlib.pyplot as plt

colorlist = ([(0.04, 0.015, 0.02, 1.0)] * 1 + [(0.8, 0.5, 0.2, 1.0)] * 1) * 16

fig, ax = plt.subplots()
y = [1] * 32
x = range(32)
ax.bar(x, y, color=colorlist)
plt.show()

However, if preferred, it also works with the np.array and either an object or tuple dtype, as well as with the given numpy data types. I just wanted to show the result can be achieved with basic Python and matplotlib alone.

edited Sep 14, 2021 at 12:22

answered Sep 13, 2021 at 5:25

Grismar

32.4k6 gold badges43 silver badges69 bronze badges

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

5 Comments

ShaunH Over a year ago

Thank you Grismar, this is nice and simple and clear. It does exactly what you said it does. Unfortunatly it doesn't work in the ax.bar( statement, says "ValueError: Invalid RGBA argument:" presumably because it's not in a numpy array with dtype=object?

Grismar Over a year ago

I've added a suggestion to cast accordingly; do note that this seems to translate the inner tuples into lists / arrays as well - but I expect that shouldn't be an issue

ShaunH Over a year ago

Unfortunately the translation from tuple to lists does make a difference to ax.bar. Some error, "Invalid RGBA argument". It's weird, why does it change it from a tuple to a list/array?

Grismar Over a year ago

If you provide a more complete example in your code, that you would expect to work, I'd be happy to give it a try

ShaunH Over a year ago

Thanks Grismar, I appreciate it. I've edited the original question to include an example. It doesn't work for me, presumably because the RGBA values are tuples.

Collectives™ on Stack Overflow

Numpy array of alternating tuples

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related