Color ramps
--colorize accepts 36 ramps from freely licensed sources: matplotlib, seaborn, cmocean, Fabio Crameri’s Scientific Colour Maps and ColorBrewer (see THIRD-PARTY-LICENSES.md for attributions). Good starting points: viridis for elevation, turbo for intensity, rdbu with stretch=minmax for change detection, las-classification for class codes. Colorizing points explains how to apply them.
Perceptually uniform
Section titled “Perceptually uniform”The safest default choices: equal steps in the data look like equal steps in color, so the ramp adds no bands or edges of its own.
| Ramp | Preview | Best for |
|---|---|---|
viridis |
General-purpose elevation or continuous density; the scientific standard. | |
viridis-pastel |
A soft, pastel take on viridis that keeps its perceptual ordering. | |
magmainfernoplasma |
Heat-like sequential data with a dark-to-bright look. | |
cividis |
Optimized for color-vision deficiency. | |
turbo |
Vibrant rainbow alternative to “jet” with tuned lightness; great for intensity and edge spotting. | |
batlow |
Colorblind-safe multi-hue rainbow alternative (Crameri); ideal for canopy-height models. |
Terrain and elevation
Section titled “Terrain and elevation”| Ramp | Preview | Best for |
|---|---|---|
terrain |
Classic land elevation: blue lowlands through green, yellow and brown to white peaks. | |
gist-earth |
Like terrain with deeper earth tones; pairs well with hillshading. | |
topo |
Combined bathymetry + land elevation (cmocean); dark depths to bright uplands. | |
oleron |
Perceptually uniform olive-to-brown ramp for bare-earth DEMs and dryland topography. | |
nuuk |
Deep indigo to bright mint; crisp accents for urban building heights. |
Sequential intensity and density
Section titled “Sequential intensity and density”Single direction, dark to bright.
| Ramp | Preview | Best for |
|---|---|---|
grayscaleheat |
Simple built-in defaults. | |
cubehelix |
Monotonically increasing brightness; stays readable in black-and-white prints. | |
mako |
Dark navy to light turquoise; deep-water and coastal bathymetric LiDAR. | |
rocket |
Blackish-purple through reds to pale cream; point-density heat maps. | |
haline |
Deep blue to bright green (cmocean); brackish water and estuaries. | |
amp |
Light-to-deep red (cmocean); laser/radar intensity over dark basemaps. | |
ylgnbu |
ColorBrewer Yellow-Green-Blue; density distributions and drainage. | |
blues |
ColorBrewer single-hue blue; water depth, subtle underlays. |
Diverging change detection
Section titled “Diverging change detection”Pair them with stretch=minmax or symmetric data, so the midpoint lands on your baseline.
| Ramp | Preview | Best for |
|---|---|---|
rdbu |
The gold standard for change: red = loss/erosion, blue = gain/accumulation. | |
brbg |
Brown-to-blue-green; bare soil versus vegetation/water shifts. | |
coolwarm |
Balance-adjusted blue-to-red that tolerates hillshading and shadows. | |
spectral |
Full-spectrum diverging ramp for deviations from a baseline. | |
balance |
Perceptually uniform blue-white-red (cmocean) with a truly neutral center. | |
seismic |
High-contrast deep-blue/white/deep-red; subsurface and elevation differences. | |
roma |
Crameri diverging scheme tailored to topographic anomalies. | |
berlin |
Dark-centered diverging map for dark-mode viewers. | |
piyg |
Pink-to-yellow-green; NDVI-style vegetation anomalies. |
Categorical classification
Section titled “Categorical classification”| Ramp | Preview | Best for |
|---|---|---|
las-classification |
ASPRS LAS class codes with their conventional colors (fixed 0–255 scale). | |
dark2 |
Distinct muted dark tones; urban feature separation. | |
paired |
Light/dark hue pairs; nested classes like low vs. high vegetation. | |
set2 |
Pastel palette that keeps overlaid labels legible. | |
accent |
Makes selected categories pop against neutral basemaps. |
Categorical ramps map equal-width value bands to discrete colors: combine them with stretch=minmax (or data with a known range) so category codes land on stable bands. las-classification needs nothing extra, as it is pinned to the absolute 0–255 class range.