Cleopatra

Cleopatra is a matplotlib utility package for visualizing 2D/3D numpy arrays, unstructured meshes, point clouds, vector fields, polygons, lines, and statistical distributions. It targets scientific and research users working with geospatial and raster data, providing a high-level API over matplotlib with sensible defaults and rich customization.

For the package's boundaries — what belongs here and what does not — see SCOPE.md.

Package Layout

graph TD
    subgraph core["Core"]
        glyph["<b>glyph</b><br/>Glyph — base class<br/>figure/axes · color norms · classification<br/>colorbars · ticks · point overlays · animation"]
    end

    subgraph geomixin["Geo mixin"]
        geo["<b>geo</b><br/>GeoMixin — crs · add_tiles<br/>add_features · add_relief<br/>add_reference_map · add_labels"]
    end

    subgraph visualizers["Visualizers — subclass Glyph"]
        array_glyph["<b>array_glyph</b><br/>ArrayGlyph · FacetGrid<br/>2D/3D rasters, facets, animation"]
        mesh_glyph["<b>mesh_glyph</b><br/>MeshGlyph<br/>unstructured meshes"]
        scatter_glyph["<b>scatter_glyph</b><br/>ScatterGlyph<br/>point clouds"]
        vector_glyph["<b>vector_glyph</b><br/>VectorGlyph<br/>vector fields"]
        flow_glyph["<b>flow_glyph</b><br/>FlowGlyph<br/>flow paths"]
        line_glyph["<b>line_glyph</b><br/>LineGlyph<br/>line / bar / band"]
        polygon_glyph["<b>polygon_glyph</b><br/>PolygonGlyph<br/>polygon collections"]
        kde_glyph["<b>kde_glyph</b><br/>KDEGlyph<br/>2D kernel density"]
    end

    subgraph standalone["Standalone"]
        histogram_glyph["<b>histogram_glyph</b><br/>HistogramGlyph<br/>histogram · boxplot · multiboxplot · stripes"]
    end

    subgraph support["Supporting utilities"]
        styles["<b>styles</b><br/>Styles · Scale · ColorScale<br/>MidpointNormalize · classify · resolve_sizes · legends"]
        colors["<b>colors</b><br/>Colors · haze data styles<br/>hex/RGB · colormaps · alpha-scaled layers"]
        animation["<b>animation</b><br/>save_animation · to_gif/mp4 · embed_gif<br/>GIF/WebP/MP4/MOV/AVI · bundled ffmpeg"]
        projection["<b>projection</b><br/>apply_projection_frame<br/>orthographic globe presets"]
        config["<b>config</b><br/>Config — matplotlib backend helper"]
    end

    subgraph optional["Optional — cleopatra[tiles]"]
        tiles["<b>tiles</b><br/>add_tiles · fetch / stitch helpers<br/>XYZ web-tile basemaps"]
        reference["<b>reference</b><br/>add_features · add_relief<br/>Natural Earth · hypsometric relief"]
    end

    array_glyph & mesh_glyph & scatter_glyph & vector_glyph & flow_glyph & line_glyph & polygon_glyph & kde_glyph ==>|extends| glyph
    array_glyph & mesh_glyph & scatter_glyph & vector_glyph & flow_glyph & polygon_glyph -.->|mixes in| geo
    geo -->|basemap tiles| tiles
    geo -->|coastlines · relief| reference
    glyph -->|color scales · classification| styles
    glyph -->|save / embed| animation

• glyph provides the shared Glyph base class (figure/axes lifecycle, colorbars, color norms, ticks, classification, animation).
• The user-facing visualizers all subclass Glyph and share its colour-mapping/colorbar pipeline — array_glyph (ArrayGlyph, FacetGrid), mesh_glyph (MeshGlyph), scatter_glyph (ScatterGlyph), vector_glyph (VectorGlyph), flow_glyph (FlowGlyph), line_glyph (LineGlyph), polygon_glyph (PolygonGlyph), and kde_glyph (KDEGlyph). histogram_glyph (HistogramGlyph) stands alone.
• geo provides GeoMixin, mixed into the six geographic visualizers — array_glyph, mesh_glyph, scatter_glyph, vector_glyph, flow_glyph, and polygon_glyph (not line_glyph, kde_glyph, or histogram_glyph) — adding a settable crs plus one-call basemap helpers on the glyph's own axes: add_tiles, add_features, add_relief, add_reference_map, and add_labels.
• tiles and reference are the optional (cleopatra[tiles]) basemap data sources geo wraps — tiles fetches/stitches XYZ web-tile mosaics, reference draws fixed public Natural Earth vector layers and a hypsometric relief raster.
• colors, styles, animation, projection, and config are supporting utilities (colour conversions plus composable "haze"-style data layers via apply_data_style and alpha-scaled image/mesh rendering; predefined styles, MidpointNormalize, ColorScale, value→size mapping, classify classification schemes and legend builders; glyph-independent animation save/embed helpers spanning GIF/WebP/MP4/MOV/AVI with a bundled-ffmpeg fallback; static projected map frames plus orthographic globe reprojection presets; and the matplotlib-backend helper).

Main Features

ArrayGlyph -- Raster / Array Visualization

• Plot 2D numpy arrays with automatic colorbar and customizable color scales (linear, power, symmetric log-norm, boundary-norm, midpoint).
• Display cell values and overlay point markers on the plot.
• Animate 3D single-band stacks or 4D RGB/RGBA true-colour stacks over time, and export to GIF, WebP, MP4, MOV, or AVI (bundled ffmpeg -- no separate install needed).
• Drop in a CAMS-style basemap (coastlines, borders, graticule) with a single add_reference_map call.

MeshGlyph -- Unstructured Mesh Visualization

• Visualize UGRID-style unstructured mesh data using triangulation (tripcolor, tricontourf).
• Render wireframe outlines via LineCollection.
• Accepts raw numpy arrays of node coordinates and face-node connectivity.
• Animate time-varying mesh data.

HistogramGlyph -- Distribution Plots

• Create histograms for 1D and 2D datasets with customizable bins, colors, and transparency.
• Draw boxplots, multi-boxplots, and strip plots.

ScatterGlyph -- Point Clouds

• Plot 2D point clouds, colour-mapped by a per-point values array with a matching colorbar.
• Encode a second quantity through per-point marker sizes (with an optional size legend), so colour and size carry two variables at once.

VectorGlyph -- Vector Fields

• Render 2D (u, v) vector fields as arrows (quiver), wind barbs, or streamlines.
• Colour the artist by vector magnitude hypot(u, v) through the shared scalar-mapping pipeline.

FlowGlyph -- Flow Paths

• Draw a sequence of polylines as a LineCollection, colour-mapped by a per-path values array.
• Scale per-path line widths by magnitude, with an optional width legend.

LineGlyph -- Line / Bar / Band Plots

• Line, bar, and fill_between (band) plots. line accepts 1D or 2D y (one series per column); bar takes a single 1D series.

PolygonGlyph -- Polygon Collections

• Fill and colour-map collections of polygons by a per-polygon values array, or draw outlines only.

KDEGlyph -- Kernel Density

• Estimate a 2D Gaussian kernel density of an (x, y) point cloud (NumPy only, no scipy) and draw it as filled or line density contours.

Geospatial basemaps -- GeoMixin

• ArrayGlyph, MeshGlyph, ScatterGlyph, VectorGlyph, FlowGlyph, and PolygonGlyph mix in GeoMixin, adding a settable crs plus one-call basemap helpers on glyph.ax: add_tiles (XYZ web-tile mosaics), add_features / add_relief (Natural Earth coastlines, borders, land, ocean, rivers, lakes, and a hypsometric relief backdrop), a one-call add_reference_map preset ("light", "dark", or "auto"), and add_labels for city/point labels.
• tiles and the fixed-public-dataset reference layers require the cleopatra[tiles] extra.

Composable data styles & globe projections

• colors.apply_data_style renders one or more layers with a named preset (currently "haze", an aerosol / organic-matter / dust look) -- per-pixel opacity tied to value via alpha_scaled_image / alpha_scaled_mesh, plus a swatch legend, in one call.
• projection.apply_projection_style reprojects (lon, lat, data) onto an orthographic "globe" view (or leaves it flat) via named presets, pairing with apply_data_style to build CAMS-style globe animations in a few lines. The orthographic helpers require the cleopatra[tiles] extra (pyproj).

Colors -- Color Utilities

• Convert between hex, RGB (0-255), and normalized RGB (0-1) formats.
• Extract color ramps from images and create custom matplotlib colormaps.
• Ready-made "haze" colormaps and alpha-scaled rendering helpers for the composable data styles above.

Styling with grouped options

• Every glyph's plot() / animate() takes small, discoverable typed objects instead of a long list of loose keyword arguments: color=ColorScaling(...), contour=Contour(levels=...), cells=CellValues(...), classify=Classify(...), data_style=DataStyle(style=..., hillshade=...), and colorbar=ColorBar(...).
• ArrayGlyph adds points=PointOverlay(...), frame_label=FrameLabel(...), facet(labels=PanelLabels(...)), and ArrayGlyph(array, rgb_bands=RgbBands([r, g, b], surface_reflectance=..., percentile=...)) for RGB composites.
• Migrating from the old loose-keyword API (e.g. rgb=, cutoff=, col_coords=, text_colors=)? See the migration guide.

Installation

pip

pip install cleopatra

# with the optional web-tile basemap support (cleopatra.basemap.tiles.add_tiles)
pip install "cleopatra[tiles]"

conda

conda install -c conda-forge cleopatra

# with the optional web-tile basemap support
conda install -c conda-forge cleopatra-tiles

The conda packages are built from the cleopatra-feedstock (the cleopatra-tiles output bundles pillow, pyproj, and xyzservices).

From source (latest development version)

pip install git+https://github.com/serapeum-org/cleopatra

Quick Start

Plot a 2D array

import numpy as np
from cleopatra.glyphs.gridded.array_glyph import ArrayGlyph

arr = np.random.rand(10, 10)
glyph = ArrayGlyph(arr)
fig, ax = glyph.plot(title="Random Array")

Create a histogram

import numpy as np
from cleopatra.glyphs.stats.histogram_glyph import HistogramGlyph

data = np.random.normal(0, 1, 1000)
stat = HistogramGlyph(data)
fig, ax = stat.histogram(bins=30)

Plot an unstructured mesh

import numpy as np
from cleopatra.glyphs.gridded.mesh_glyph import MeshGlyph

node_x = np.array([0.0, 1.0, 0.5, 1.5])
node_y = np.array([0.0, 0.0, 1.0, 1.0])
face_nodes = np.array([[0, 1, 2], [1, 3, 2]])
face_data = np.array([10.0, 20.0])

mg = MeshGlyph(node_x, node_y, face_nodes)
fig, ax = mg.plot(face_data, location="face", title="Mesh Data")

Plot a value-coloured point cloud

import numpy as np
from cleopatra.glyphs.primitives.scatter_glyph import ScatterGlyph

x = np.random.rand(100)
y = np.random.rand(100)
values = np.random.rand(100)
sg = ScatterGlyph(x, y, values=values)
fig, ax, sc = sg.plot(title="Scatter")

Plot a vector field

import numpy as np
from cleopatra.glyphs.gridded.vector_glyph import VectorGlyph

x, y = np.meshgrid(np.linspace(0, 1, 8), np.linspace(0, 1, 8))
u, v = np.cos(x), np.sin(y)
vg = VectorGlyph(x, y, u, v)
fig, ax, artist = vg.plot(kind="quiver", title="Vector Field")

Add a basemap to an array plot

import numpy as np
from cleopatra.glyphs.gridded.array_glyph import ArrayGlyph

field = np.random.rand(80, 120)
glyph = ArrayGlyph(field, extent=[-100, 15, -40, 55])  # west, south, east, north
glyph.plot(cmap="turbo", cbar_label="anomaly")
glyph.add_reference_map("light")  # coastlines, borders, and a lon/lat graticule

Requirements

• Python >= 3.11
• numpy >= 2.0.0
• matplotlib >= 3.9

Ships with a bundled ffmpeg binary (via imageio-ffmpeg), so save_animation can export MP4/MOV/AVI without a separate system install. Geospatial basemaps and globe-projection presets (GeoMixin, cleopatra.basemap.tiles, cleopatra.basemap.reference, and the orthographic helpers in cleopatra.basemap.projection) need the cleopatra[tiles] extra.

Documentation

Full documentation is available at serapeum-org.github.io/cleopatra. Upgrading across a breaking release? See the migration guide.

License

Cleopatra is licensed under the GNU General Public License v3.