Simple heatmap

In [1]:

from plotnine import (
    ggplot,
    aes,
    theme_matplotlib,
    theme_set,
    geom_tile,
    scale_fill_continuous,
    coord_cartesian
)

from plotnine.data import faithfuld

# Set default theme for all the plots
theme_set(theme_matplotlib())

Map data to color gradient

Without specifically calling scale_fill_continuous(), a default scale is applied to geom_tile().

In [2]:

faithfuld.head()

	eruptions	waiting	density
0	1.600000	43.0	0.003216
1	1.647297	43.0	0.003835
2	1.694595	43.0	0.004436
3	1.741892	43.0	0.004978
4	1.789189	43.0	0.005424

In [3]:

(
    ggplot(faithfuld, aes("waiting", "eruptions", fill="density")) 
    + geom_tile()
)

Notice that adding scale_fill_continuous() will not change the default plot.

In [4]:

(
    ggplot(faithfuld, aes("waiting", "eruptions", fill="density"))
    + geom_tile()
    + scale_fill_continuous()
)

The scale aesthetics can be edited to other color maps. Use the cmap_name argument to specify an alternative palette. Find other available color maps in the matplotlib documentation.

In [5]:

# Gallery, tiles

(
    ggplot(faithfuld, aes("waiting", "eruptions", fill="density"))
    + geom_tile()
    + scale_fill_continuous(cmap_name="plasma")
    + coord_cartesian(expand=False)
)