I have a problem concerning the management of a colormap.
In the figure below, each point displayed is associated with a value : 2, 3, 4, 5 or 13.
So, to know the value associated with each point, I used the colormap option. I generated a colormap of 5 colors, one color associated with each value.
The problem is that the values 2, 3 and 4 are associated with purple, the value 5 with blue and the value 13 with red. Indeed, the colormap contains 5 colors, but the scale of the labels on the right goes from 2 to 13.
However, I would like the 2 to be associated with purple, the 3 with blue, the 4 with green, the 5 with orange and the 13 with red.
My code:
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Sat May 14 15:04:21 2022
@author: cldelh
"""
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.pyplot import cm
# Abscisses
W = np.array([0. , 0.01 , 0.01999402, 0.03412597, 0.05410511,
0.08232443, 0.12194226, 0.16661301, 0.20758763, 0.23904194,
0.24626125, 0.25039445, 0.25330903, 0.25582872, 0.25816522,
0.26031379, 0.26266544, 0.26568211, 0.27005595, 0.2770875 ,
0.28909234, 0.30636753, 0.32544111, 0.34694002, 0.36917884,
0.3897257 , 0.40656458, 0.41784063, 0.42541583, 0.43055847,
0.43452519, 0.43742307, 0.43919769, 0.44031273, 0.44112462,
0.44200454, 0.44347889, 0.44643343, 0.45229159, 0.46288266,
0.48056282, 0.50895736, 0.54927625, 0.59608409, 0.64487664,
0.7016973 , 0.76708481, 0.84185322, 0.92696579, 1.01707062,
1.10557259, 1.19256251, 1.27815237, 1.36244217, 1.44550112,
1.52735963, 1.60800828, 1.68740049, 1.76545689, 1.84207015,
1.91710981, 1.99042675, 2.06185747, 2.13122796, 2.1983573 ,
2.263061 , 2.32515405, 2.38445365, 2.44078163, 2.49396659,
2.54384564, 2.5902658 , 2.63308518, 2.67217368, 2.70741361,
2.73869997, 2.76594058, 2.78905614, 2.80798011, 2.82265863,
2.83305046, 2.83912688, 2.84087179, 2.83828174, 2.83136621,
2.82014797, 2.80466344, 2.78496333, 2.76111323, 2.73319431,
2.70130406, 2.665557 , 2.62608541, 2.58303999, 2.53659044,
2.48692603, 2.43425621, 2.3788113 , 2.32084358, 2.26062905,
2.1984705 , 2.13470266, 2.06970116, 2.0038978 , 1.93780705,
1.87207326, 1.80755842, 1.74551542, 1.68795847, 1.63852865,
1.60461127, 1.60094385, 1.64165693, 1.71498642, 1.80255956,
1.89604738, 1.99229146, 2.08996895, 2.18845668, 2.28743203,
2.38671536, 2.48620089, 2.58582339, 2.68554117, 2.78532671,
2.8851614 , 2.98503236, 3.08493046])
# Ordonnées
Xinf = np.array([0.36001558, 0.36002501, 0.36005669, 0.36013731, 0.36031898,
0.36062301, 0.36527713, 0.36485195, 0.38101038, 0.39612248,
0.40736786, 0.41642505, 0.42372367, 0.4296357 , 0.4329689 ,
0.43243038, 0.42777775, 0.41881406, 0.4058955 , 0.39050351,
0.37868412, 0.37464251, 0.3753022 , 0.38049347, 0.39215149,
0.41261855, 0.44186649, 0.47214887, 0.50022204, 0.52476197,
0.54846522, 0.57015033, 0.58647846, 0.59804236, 0.60505114,
0.60736793, 0.60431747, 0.59481333, 0.57794772, 0.55477589,
0.52875494, 0.50564226, 0.49269248, 0.49103505, 0.49670408,
0.50868991, 0.52725566, 0.55320582, 0.58777717, 0.62935965,
0.67447035, 0.72228144, 0.77209809, 0.82333207, 0.87550968,
0.92821221, 0.98109947, 1.03388713, 1.08631077, 1.13814125,
1.18917466, 1.23921078, 1.28807521, 1.33558548, 1.38158777,
1.42589996, 1.46839252, 1.50890861, 1.54730779, 1.58347723,
1.61727155, 1.64859434, 1.67732612, 1.70338193, 1.72667917,
1.74713082, 1.76466827, 1.77924246, 1.79079041, 1.79926455,
1.80462534, 1.80686203, 1.80593978, 1.80183511, 1.79455897,
1.78410278, 1.77046371, 1.75366114, 1.733729 , 1.7106789 ,
1.68454743, 1.65537695, 1.62321454, 1.58811057, 1.55013282,
1.50933504, 1.46577577, 1.41951312, 1.37061227, 1.31911356,
1.2650476 , 1.20840623, 1.14917691, 1.08726763, 1.02253009,
0.95471075, 0.88341169, 0.8079946 , 0.72748982, 0.64047401,
0.54539278, 0.44410944, 0.35174988, 0.28325452, 0.23475268,
0.19915439, 0.17195691, 0.15050578, 0.13316766, 0.11888317,
0.10692915, 0.09679719, 0.08811551, 0.08060803, 0.07406247,
0.06831513, 0.06323702, 0.05872426])
# Variables
variable = np.array([ 5., 2., 2., 2., 2., 2., 3., 4., 5., 5., 13., 4., 3.,
3., 4., 4., 4., 4., 4., 4., 4., 4., 3., 3., 3., 3.,
4., 4., 4., 3., 3., 4., 4., 4., 4., 4., 4., 4., 4.,
4., 4., 4., 4., 4., 3., 3., 3., 3., 3., 3., 3., 3.,
3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3.,
3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3.,
3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3.,
3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3., 3.,
3., 3., 3., 3., 3., 3., 3., 4., 4., 4., 4., 3., 3.,
3., 3., 2., 2., 2., 2., 2., 2., 2., 2., 2.])
# Valeurs contenues dans variable
variable_delete = np.array([ 5., 2., 3., 4., 13.])
# Points
plt.scatter(W,Xinf,s=10,marker='o', c=variable, cmap=cm.get_cmap('rainbow', len(variable_delete)))
# Colorbar
cb = plt.colorbar(ticks=variable_delete, location='right', label=r'$k_{iter} \; [\emptyset]$')
CodePudding user response：
The following approach first locates 5 equally spaced spots into the range from 0 to 1 (the default colormap range). Then it maps each of the 5 variable_delete values to its corresponding spot. These mapped values are used for the coloring. Afterwards, the colorbar positions get labeled accordingly.
The code also works when the numeric labels are in a random order.
(An alternative could be to work with a BoundaryNorm, but that gets more complicated to nicely locate the labels.)
import numpy as np
import matplotlib.pyplot as plt
# W = ...
# Xinf = ...
# variable = ...
variable_delete = np.array([2., 3., 4., 5., 13.])
cbar_spots = np.linspace(0, 1, 2 * len(variable_delete) 1)[1::2]
cbar_dict = {var: spot for var, spot in zip(variable_delete, cbar_spots)}
var_mapped = np.zeros_like(variable)
for k, v in cbar_dict.items():
var_mapped[variable == k] = v
plt.scatter(W, Xinf, s=10, marker='o', c=var_mapped,
cmap=plt.get_cmap('rainbow', len(variable_delete)), vmin=0, vmax=1)
cb = plt.colorbar(ticks=cbar_spots, location='right', label=r'$k_{iter} \; [\emptyset]$')
cb.set_ticklabels(variable_delete.astype(int).astype(str))
plt.show()
CodePudding user response：
May I suggest a simple alternative to using a color map?
In [26]: for v in sorted(variable_delete):
...: ix = variable == v
...: plt.scatter(W[ix], Xinf[ix], s=6, marker='o', label='%3.1f'%v)
...: plt.legend()
...: plt.grid()
