Plot RGB map from three columns of data using Matplotlib

Question

Is it possible to plot a map of RGB values using Matplotlib?

I have three columns of data that I read from a text file in the following form where x and y are the desired coordinates and z is a hex string of the desired rgb color to plot at the given coordinates:

x   y   z 

1  0.5  #000000 

2  0.5  #FF0000 

3  0.5  #00FF00 

1  1.5  #0000FF 

2  1.5  #FFFF00 

3  1.5  #00FFFF 

1  1.5  #FF00FF 

2  2.5  #C0C0C0 

3  2.5  #FFFFFF

This is my current state of code. An error is thrown from the griddata() function:

import pandas as pds
import matplotlib.pyplot as plt

# Import text file using pandas
datafile = pds.read_csv(PathToData,sep='\t')

X=datafile.x
Y=datafile.y
Z=datafile.z

# Generate mesh as 'numpy.ndarray' type for plotting
# This throws the following error:
# ValueError: could not convert string to float: #FFAA39
Z=griddata(X, Y, Z, unique(X), unique(Y))

Many thanks

I have been experimenting with pcolor() and imshow() but neither seem to be working correctly for me at the moment. Thanks — atstack
– atstack, Commented Mar 14, 2013 at 0:15
You should post your attempts. People are much more willing to help you fix code you already have than flat out write code for you. Is your problem you can't parse the text file? If not, we don't need to know you are importing from a text file once you have the data in arrays. Is imshow giving you errors? If so what errors? If not, what is it doing (or not doing) that is different from your expectations? — tacaswell
– tacaswell, Commented Mar 14, 2013 at 1:10

tacaswell · Accepted Answer · 2013-03-14 12:54:16Z

1

griddata is a function for interpolating unevenly spaced data on to a grid (griddata doc). In your case, it looks like you already have data on a grid, you just need to reshape it. The error you are getting arises because griddata is trying to convert you color hex codes to floats for the interpolation, which you should be getting because there is not a sensible float interpertation of #00FF00.

data_dims = (3, 3) # or what ever your data really is
X = np.asarray(x).reshape(data_dims)
Y = np.asarray(y).reshape(data_dims)
C = np.asarray([mpl.colors.colorConverter.to_rgb(_hz) for _hz in z]).reshape(data_dims + (3,)) 
# the colors converted to (r, g, b) tuples

The asarray calls are to make sure we have arrays, not data frames.

If you just want to see the array we can use imshow to plot it:

imshow(c, interpolation='none', 
       origin='bottom', 
       extent=[np.min(X), np.max(X), np.min(Y), np.max(Y)])

color converter doc, imshow doc.

You might need to transpose/flip/rotate C to get the orientation you expect.

answered Mar 14, 2013 at 12:54

tacaswell

88.2k23 gold badges222 silver badges203 bronze badges

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

2 Comments

atstack Over a year ago

Thanks. That is almost perfect, however, axis numbering is a little bit off... is there a simple way to fix that? Here is a link to the output link

tacaswell Over a year ago

play with the values in extent. That box is the box that the result will fill.

Collectives™ on Stack Overflow

Plot RGB map from three columns of data using Matplotlib

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