Plots and charts

Non-map figures use fastplot.pyplot, which follows matplotlib's pyplot: the current figure and axes, plt.subplots, every Axes method listed below, rcParams and the style sheets. They are drawn the way matplotlib's Agg backend draws: lines snapped to pixels, the same caps and joins, the same path simplification, ticks, labels and layout.

import fastplot.pyplot as plt

fig, ax = plt.subplots(figsize=(7, 4), dpi=120)
ax.plot(x, y, 'o-', label='observed')
ax.set_xlabel('Distance (km)')
ax.legend()
fig.tight_layout()
fig.savefig('profile.png')

What you can draw

Kind	Methods
Lines and points	plot (format strings, every marker), scatter, step, stairs, stem, errorbar, hlines, vlines, axhline, axvline
Areas	fill_between, fill_betweenx (with where, interpolate, step), stackplot, axhspan, axvspan
Bars and counts	bar, barh (error bars, bar_label), hist (bar, step, stepfilled; stacked, density, cumulative)
Fields	pcolormesh, imshow, contourf, contour, clabel
Vectors	quiver, quiverkey, barbs, streamplot
Triangulations	tricontourf, tricontour, tripcolor, triplot, with fastplot.tri
Shapes	add_patch with Rectangle, Circle, Ellipse, Polygon, Wedge, Arc, RegularPolygon, FancyBboxPatch, FancyArrowPatch
Text	text, annotate (every arrow and connection style), legend, fig.legend, suptitle

Hatching works on bars, fills, patches and contourf(hatches=...).

Axes and scales

• Linear, log, symlog, logit and function scales, with semilogx, semilogy and loglog.
• Dates from datetime, numpy.datetime64 or pandas, with fastplot.dates locators and formatters.
• Categories from strings.
• twinx, twiny, sharex, sharey, inverted axes, aspect ratios, margins and sticky edges.
• fastplot.ticker locators and formatters. matplotlib's own ticker and date objects also work.
• Spines, tick_params, minor ticks and grids.

Layout

subplots, subplot_mosaic, GridSpec, add_axes, subplots_adjust and tight_layout place axes with matplotlib's arithmetic, so a figure keeps its proportions when you switch libraries.

Polar axes

fig.add_subplot(projection='polar') gives matplotlib's polar axes for the full circle. set_theta_zero_location, set_theta_direction, set_rlim, set_rorigin, set_rlabel_position, set_thetagrids and set_rgrids are all there, and a log radius works. Bars on polar axes make wind roses; pcolormesh and contourf draw radar sweeps.

Triangulations

fastplot.tri matches matplotlib.tri: Triangulation, TriFinder, LinearTriInterpolator, CubicTriInterpolator, UniformTriRefiner and TriAnalyzer. With scipy installed, triangulations are the same triangles in the same order as matplotlib's. Without it, fastplot uses its own Delaunay triangulation.

Style

plt.style.use('ggplot')
with plt.rc_context({'font.size': 9, 'axes.grid': True}):
    fig, ax = plt.subplots()

All 31 of matplotlib's style sheets ship with fastplot, and rcParams takes the same keys.

Colours

Every matplotlib colormap and named colour, ListedColormap, LinearSegmentedColormap.from_list, and the norms Normalize, LogNorm, BoundaryNorm, TwoSlopeNorm, CenteredNorm, SymLogNorm and NoNorm. extend and the under, over and bad colours work as in matplotlib.