| 1 |
""" |
| 2 |
.. _colormap_example: |
| 3 |
|
| 4 |
Colormap Choices |
| 5 |
~~~~~~~~~~~~~~~~ |
| 6 |
|
| 7 |
Use a Matplotlib, Colorcet, cmocean, or custom colormap when plotting scalar |
| 8 |
values with :func:`pyvista.plot` and :class:`~pyvista.Plotter` methods. |
| 9 |
""" |
| 10 |
|
| 11 |
from matplotlib.colors import LinearSegmentedColormap |
| 12 |
from matplotlib.colors import ListedColormap |
| 13 |
import matplotlib.pyplot as plt |
| 14 |
import numpy as np |
| 15 |
import pyvista as pv |
| 16 |
from pyvista import examples |
| 17 |
|
| 18 |
# %% |
| 19 |
# Any colormap built for ``matplotlib``, ``colorcet``, or ``cmocean`` is fully |
| 20 |
# compatible with PyVista. Colormaps are typically specified by passing the |
| 21 |
# string name of the colormap to the plotting routine via the ``cmap`` |
| 22 |
# argument. |
| 23 |
# |
| 24 |
# See `Matplotlib's complete list of available colormaps`_, |
| 25 |
# `Colorcet's complete list`_, and `cmocean's complete list`_. |
| 26 |
# |
| 27 |
# .. _Matplotlib's complete list of available colormaps: https://matplotlib.org/tutorials/colors/colormaps.html |
| 28 |
# .. _Colorcet's complete list: https://colorcet.holoviz.org/user_guide/index.html |
| 29 |
# .. _cmocean's complete list: https://matplotlib.org/cmocean/ |
| 30 |
|
| 31 |
# %% |
| 32 |
# Custom Made Colormaps |
| 33 |
# +++++++++++++++++++++ |
| 34 |
# |
| 35 |
# To get started using a custom colormap, download some data with scalar values to |
| 36 |
# plot. |
| 37 |
|
| 38 |
mesh = examples.download_st_helens().warp_by_scalar() |
| 39 |
# Add scalar array with range (0, 100) that correlates with elevation |
| 40 |
mesh['values'] = pv.plotting.tools.normalize(mesh['Elevation']) * 100 |
| 41 |
|
| 42 |
# %% |
| 43 |
# Build a custom colormap - here we make a colormap with 5 discrete colors |
| 44 |
# and we specify the ranges where those colors fall: |
| 45 |
|
| 46 |
# Define the colors we want to use |
| 47 |
blue = np.array([12 / 256, 238 / 256, 246 / 256, 1.0]) |
| 48 |
black = np.array([11 / 256, 11 / 256, 11 / 256, 1.0]) |
| 49 |
grey = np.array([189 / 256, 189 / 256, 189 / 256, 1.0]) |
| 50 |
yellow = np.array([255 / 256, 247 / 256, 0 / 256, 1.0]) |
| 51 |
red = np.array([1.0, 0.0, 0.0, 1.0]) |
| 52 |
|
| 53 |
mapping = np.linspace(mesh['values'].min(), mesh['values'].max(), 256) |
| 54 |
newcolors = np.empty((256, 4)) |
| 55 |
newcolors[mapping >= 80] = red |
| 56 |
newcolors[mapping < 80] = grey |
| 57 |
newcolors[mapping < 55] = yellow |
| 58 |
newcolors[mapping < 30] = blue |
| 59 |
newcolors[mapping < 1] = black |
| 60 |
|
| 61 |
# Make the colormap from the listed colors |
| 62 |
my_colormap = ListedColormap(newcolors) |
| 63 |
|
| 64 |
# %% |
| 65 |
# Simply pass the colormap to the plotting routine. |
| 66 |
mesh.plot(scalars='values', cmap=my_colormap) |
| 67 |
|
| 68 |
# %% |
| 69 |
# Or you could make a simple colormap... any Matplotlib colormap can be passed |
| 70 |
# to PyVista. |
| 71 |
boring_cmap = plt.get_cmap('viridis', 5) |
| 72 |
mesh.plot(scalars='values', cmap=boring_cmap) |
| 73 |
|
| 74 |
# %% |
| 75 |
# You can also pass a list of color strings to the color map. This |
| 76 |
# approach divides up the colormap into 5 equal parts. |
| 77 |
mesh.plot(scalars=mesh['values'], cmap=['black', 'blue', 'yellow', 'grey', 'red']) |
| 78 |
|
| 79 |
# %% |
| 80 |
# If you still wish to have control of the separation of values, you |
| 81 |
# can do this by creating a scalar array and passing that to the |
| 82 |
# plotter along with the colormap |
| 83 |
scalars = np.empty(mesh.n_points) |
| 84 |
scalars[mesh['values'] >= 80] = 4 # red |
| 85 |
scalars[mesh['values'] < 80] = 3 # grey |
| 86 |
scalars[mesh['values'] < 55] = 2 # yellow |
| 87 |
scalars[mesh['values'] < 30] = 1 # blue |
| 88 |
scalars[mesh['values'] < 1] = 0 # black |
| 89 |
|
| 90 |
mesh.plot(scalars=scalars, cmap=['black', 'blue', 'yellow', 'grey', 'red']) |
| 91 |
|
| 92 |
# %% |
| 93 |
# Matplotlib vs. Colorcet |
| 94 |
# +++++++++++++++++++++++ |
| 95 |
# |
| 96 |
# Let's compare Colorcet's perceptually uniform "fire" colormap to Matplotlib's |
| 97 |
# "hot" colormap much like the example on the `first page of Colorcet's docs`_. |
| 98 |
# |
| 99 |
# .. _first page of Colorcet's docs: https://colorcet.holoviz.org/index.html |
| 100 |
# |
| 101 |
# The "hot" version washes out detail at the high end, as if the image is |
| 102 |
# overexposed, while "fire" makes detail visible throughout the data range. |
| 103 |
# |
| 104 |
# Please note that in order to use Colorcet's colormaps including "fire", you |
| 105 |
# must have Colorcet installed in your Python environment: |
| 106 |
# ``pip install colorcet`` |
| 107 |
|
| 108 |
pl = pv.Plotter(shape=(2, 2), border=False) |
| 109 |
pl.subplot(0, 0) |
| 110 |
pl.add_mesh( |
| 111 |
mesh, |
| 112 |
scalars='Elevation', |
| 113 |
cmap='fire', |
| 114 |
lighting=True, |
| 115 |
scalar_bar_args={'title': 'Colorcet Fire'}, |
| 116 |
) |
| 117 |
|
| 118 |
pl.subplot(0, 1) |
| 119 |
pl.add_mesh( |
| 120 |
mesh, |
| 121 |
scalars='Elevation', |
| 122 |
cmap='fire', |
| 123 |
lighting=False, |
| 124 |
scalar_bar_args={'title': 'Colorcet Fire (No Lighting)'}, |
| 125 |
) |
| 126 |
|
| 127 |
pl.subplot(1, 0) |
| 128 |
pl.add_mesh( |
| 129 |
mesh, |
| 130 |
scalars='Elevation', |
| 131 |
cmap='hot', |
| 132 |
lighting=True, |
| 133 |
scalar_bar_args={'title': 'Matplotlib Hot'}, |
| 134 |
) |
| 135 |
|
| 136 |
pl.subplot(1, 1) |
| 137 |
pl.add_mesh( |
| 138 |
mesh, |
| 139 |
scalars='Elevation', |
| 140 |
cmap='hot', |
| 141 |
lighting=False, |
| 142 |
scalar_bar_args={'title': 'Matplotlib Hot (No Lighting)'}, |
| 143 |
) |
| 144 |
|
| 145 |
pl.show() |
| 146 |
|
| 147 |
# %% |
| 148 |
# .. _calculix_colormap_example: |
| 149 |
# |
| 150 |
# Recreating Native Solver Colormaps (CalculiX) |
| 151 |
# +++++++++++++++++++++++++++++++++++++++++++++ |
| 152 |
# Sometimes you want to process results in PyVista but keep the visual |
| 153 |
# style identical to the solver's native post-processing tool. |
| 154 |
# Below is an example of how to recreate the default ``cgx`` (CalculiX GraphiX) |
| 155 |
# colormap using Matplotlib's :class:`matplotlib.colors.LinearSegmentedColormap`. |
| 156 |
# It may be particular useful when visualizing ``.frd`` files with |
| 157 |
# :class:`~pyvista.FRDReader`. |
| 158 |
# |
| 159 |
# This specific colormap uses exact RGB nodes at specific scalar fractions |
| 160 |
# (0.0, 0.1, 0.3, 0.5, 0.8, 0.9, 1.0) to create its distinct color bands. |
| 161 |
|
| 162 |
# Define the exact RGB nodes from the cgx source code |
| 163 |
cgx_nodes = [ |
| 164 |
(0.0, (0.50, 0.0, 1.0)), # Purple |
| 165 |
(0.1, (0.00, 0.0, 0.8)), # Dark Blue |
| 166 |
(0.3, (0.00, 1.0, 0.4)), # Greenish Blue |
| 167 |
(0.5, (0.00, 0.5, 0.0)), # Dark Green |
| 168 |
(0.8, (1.00, 1.0, 0.0)), # Yellow |
| 169 |
(0.9, (1.00, 0.5, 0.0)), # Orange |
| 170 |
(1.0, (0.75, 0.0, 0.0)), # Dark Red |
| 171 |
] |
| 172 |
|
| 173 |
# Create the colormap |
| 174 |
cgx_cmap = LinearSegmentedColormap.from_list('cgx', cgx_nodes) |
| 175 |
|
| 176 |
# %% |
| 177 |
# Download a sample mesh to demonstrate the colormap |
| 178 |
mesh = examples.download_notch_stress() |
| 179 |
|
| 180 |
# %% |
| 181 |
# Plot the mesh using the custom cgx colormap |
| 182 |
cpos = pv.CameraPosition( |
| 183 |
position=(0.1249, -0.1242, 0.1698), |
| 184 |
focal_point=(0.1965, 0.0467, -0.0036), |
| 185 |
viewup=(0.2512, 0.6356, 0.7299), |
| 186 |
) |
| 187 |
|
| 188 |
pl = pv.Plotter() |
| 189 |
pl.add_mesh( |
| 190 |
mesh, |
| 191 |
cmap=cgx_cmap, |
| 192 |
n_colors=21, |
| 193 |
show_edges=True, |
| 194 |
) |
| 195 |
pl.camera_position = cpos |
| 196 |
pl.show() |
| 197 |
|
| 198 |
# %% |
| 199 |
# .. tags:: plot |