Matplotlib 离散颜色条，用于绘制不同的土地利用类别

Question

I tried to define a discrete color bar to plot different land categories using separate colors for each category. I tried with the following code. I am not getting the red/tomato color for the first chosen category ie "Urban". The color bar is starting with plum color. How can I resolve it?

My other query is can we put the color bar tick at the middle of discrete color code on the color bar?

I am not attaching my dataset here. If it's required any random array can be created. Landuse categories are nothing but we represent different categories with few specified numbers(eg: 1 for urban-city, 13 for the forest, 9 for snow etc.)

import cartopy.crs as ccrs
import matplotlib.pyplot as plt
import matplotlib.colors
from matplotlib.colors import ListedColormap
import numpy as np

col_dict = {
    1:"tomato",
    2:"plum",
    7:"lightgreen",
    11:"green",
    15:"maroon",
    16:"aqua",
    17:"navy",
    19:"goldenrod",
    22:"white",
}
labels = np.array([
    "Urban",
    "I Cropland",
    "Grassland",
    "B Deciduous",
    "Mixed Forest",
    "Water Bodies",
    "H Wetland", 
    "Barren",
    "Tundra",
])
crs = ccrs.PlateCarree()
ax = plt.subplot(111, projection=crs)
cm = ListedColormap([col_dict[x] for x in col_dict.keys()])
conf = ax.contourf(lonn, latt, lu_index, cmap=cm)

Answer 1

So you are only part way there with the colormap. You need to define levels for your contours, and you need to use a BoundaryNorm. Taking your projection out of the problem and just making equal stripes, I've made the contour levels exactly half way between your discrete values. If you don't do this, the linear interpolation between data will put the contour towards one side or another.

import matplotlib.pyplot as plt
import matplotlib.colors as mcolors
from matplotlib.colors import ListedColormap
import numpy as np

col_dict = {
    1:"tomato",
    2:"plum",
    7:"lightgreen",
    11:"green",
    15:"maroon",
    16:"aqua",
    17:"navy",
    19:"goldenrod",
    22:"white",
}
labels = np.array([
    "Urban",
    "I Cropland",
    "Grassland",
    "B Deciduous",
    "Mixed Forest",
    "Water Bodies",
    "H Wetland", 
    "Barren",
    "Tundra",
])
cm = ListedColormap([col_dict[x] for x in col_dict.keys()])
np.random.seed(2112)
lu_index = np.zeros((9, 9))
for nn,k in enumerate(col_dict.keys()):
    lu_index[nn, :] = k

bounds = np.array([k for k in col_dict.keys()])
bounds = bounds[:-1] + np.diff(bounds) / 2
bounds = np.concatenate((np.array([0]), bounds, np.array([24])))
norm = mcolors.BoundaryNorm(bounds, 9)

fig, axs = plt.subplots(1, 3, constrained_layout=True, figsize=(6, 4))
ax = axs[0]
conf = ax.contourf(lu_index, cmap=cm)
ax.set_title('No Boundary Norm, \nno levels')

ax = axs[1]
conf = ax.contourf(lu_index, cmap=cm, levels=bounds)
ax.set_title('No Boundary Norm, \nlevels')

ax = axs[2]
conf = ax.contourf(lu_index, cmap=cm, norm=norm, levels=bounds)
ax.set_title('Boundary Norm, \nlevels')