I'm using matplotlib to plot a simple graph:
cm=plt.get_cmap('Blues')
nx.draw_circular(G,
node_color='White',
edge_color=range(G.number_of_edges()),
edge_cmap=cm,
node_size=900,
width=4
)
I want to set a range on the colormap 'Blues' in such a way to delete the white color which is not visible in the draw.
Please help!
Sorry for bad english.
The range (or normilization) is not really a feature of the colormap, but is often implemented as a feature in the functions that plot using colormaps. For example, imshow uses vmin and vmax, so you might try using these as keywords with draw_circular (I can't find the documentation), or maybe norm.
Other than this, you can make your own colormap with exact color arrangement that you want. There are plenty of examples on how to make custom colormaps, and a few different approaches available. Here (a, b, c, d) are a few examples that might be useful to you.
I ran into this problem trying to plot data with different colormaps:
It's hard to which of the whitish dots belong to which distribution. I solved this problem by chopping off the whiter parts of the colormap spectrum:
import matplotlib.pyplot as plt
import numpy as np
from matplotlib.colors import LinearSegmentedColormap
def chop_cmap_frac(cmap: LinearSegmentedColormap, frac: float) -> LinearSegmentedColormap:
"""Chops off the beginning `frac` fraction of a colormap."""
cmap_as_array = cmap(np.arange(256))
cmap_as_array = cmap_as_array[int(frac * len(cmap_as_array)):]
return LinearSegmentedColormap.from_list(cmap.name + f"_frac{frac}", cmap_as_array)
cmap1 = plt.get_cmap('Reds')
cmap2 = plt.get_cmap('Blues')
cmap1 = chop_cmap_frac(cmap1, 0.4)
cmap2 = chop_cmap_frac(cmap2, 0.4)
np.random.seed(42)
n = 50
xs = np.random.normal(size=n)
ys = np.random.normal(size=n)
vals = np.random.uniform(size=n)
plt.scatter(xs, ys, c=vals, cmap=cmap1)
plt.scatter(ys, xs, c=vals, cmap=cmap2)
plt.gca().set_facecolor('black')
plt.colorbar()
plt.show()
