NumPy error: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all() when not using Boolean logic

Question

I'm trying to convert a 2d NumPy array of complex numbers into another 2d array RGB values for an image, by mapping a function over the array.

This is the function:

import numpy as np
import colorsys

def ComplexToRGB(complex):
    angle = ((np.angle(complex,deg=True)+360)%360)/360
    magnitude = np.absolute(complex)
    rawColor =  colorsys.hls_to_rgb(angle,0.5,1-np.log2(magnitude)%1)
    return rawColor

I map this function over an array by passing the array as input:

print(ComplexToRGB(foo))

where foo is foo=np.linspace(-1-1j,1+1j,11).

When I pass a single complex number as an input, I don't get an error, but when I try to use an array as an input, I get the rather mysterious error

Exception has occurred: ValueError
The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

  File "C:\Users\trevo\Documents\GitHub\NumPy-Complex-Function-Grapher\colorTest.py", line 7, in ComplexToRGB
    rawColor =  colorsys.hls_to_rgb(angle,0.5,1-np.log2(magnitude)%1)
  File "C:\Users\trevo\Documents\GitHub\NumPy-Complex-Function-Grapher\colorTest.py", line 12, in <module>
    print(ComplexToRGB(foo))

I understand that you can't use Boolean conditionals when you map over an array, but I don't have any logic in my function, just math, so I'm not user where the error is coming from. Thanks for your help!

Full traceback please!

hpaulj
– hpaulj

2021-12-25 03:25:00 +00:00
Commented Dec 25, 2021 at 3:25 — hpaulj
– hpaulj, Commented Dec 25, 2021 at 3:25

hpaulj · Accepted Answer · 2021-12-25 04:08:33Z

1

With a single complex number:

In [3]: ComplexToRGB(1+3j)
Out[3]: (0.6041685072417726, 0.6695179762781593, 0.33048202372184066)

same thing with a single element array:

In [4]: ComplexToRGB(np.array(1+3j))
Out[4]: (0.6041685072417726, 0.6695179762781593, 0.33048202372184066)
In [5]: ComplexToRGB(np.array([1+3j]))
Out[5]: (array([0.60416851]), array([0.66951798]), array([0.33048202]))

but with a 2 element array:

In [6]: ComplexToRGB(np.array([1+3j]*2))
Traceback (most recent call last):
  File "<ipython-input-6-a1b8e92d5dcc>", line 1, in <module>
    ComplexToRGB(np.array([1+3j]*2))
  File "<ipython-input-1-05948df7c0bd>", line 6, in ComplexToRGB
    rawColor =  colorsys.hls_to_rgb(angle,0.5,1-np.log2(magnitude)%1)
  File "/usr/lib/python3.8/colorsys.py", line 99, in hls_to_rgb
    if s == 0.0:
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

You may not have any explicit logic in your code, but the Python colorsys code does do a if test. colorsys is a base Python module, and is not written with numpy arrays in mind.

Read the full traceback, and if necessary, read the docs of the relevant function(s).

answered Dec 25, 2021 at 4:08

hpaulj

233k14 gold badges260 silver badges392 bronze badges

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

2 Comments

Lone Survivr Over a year ago

Thanks for the explanation! Is there anything I can do to be to use it as it is, or would I have use the slower NumPy.vectorize(), or write my own color conversion?

hpaulj Over a year ago

np.vectorize is just a convenience, a list comprehension is just as good. I don't know know enough about this colorsys conversion to help with alternatives. I suppose a start would be to look at its code to see just what it is doing with the if test.

Collectives™ on Stack Overflow

NumPy error: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all() when not using Boolean logic

1 Answer 1

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related