Visualisation: Distinguishing Phases

Styling by phase is largely achieved with color, here using the phase_color() function, which will return unique colors for each phase:

from pyrolite_meltsutil.vis.style import phase_color

phase_color("olivine")

'green'

This will also work if given a phase ID:

phase_color("olivine_0")

'green'

Similarly, to differentiate between different generations or endmembers, a linestyle or marker can be used, here generated with phaseID_linestyle() which takes the full phaseID:

from pyrolite_meltsutil.vis.style import phaseID_linestyle, phaseID_marker

[phaseID_linestyle(ol) for ol in ["olivine_0", "olivine_1", "olivine_2"]]

['-', '--', ':']

[phaseID_marker(ol) for ol in ["olivine_0", "olivine_1", "olivine_2"]]

['D', 's', 'o']

We can now use these when we’re plotting to differentiate different phases:

phaseIDs = ["olivine_0", "olivine_1", "clinopyroxene_0", "spinel_0"]
styles = [
    dict(color=phase_color(ID), ls=phaseID_linestyle(ID), marker=phaseID_marker(ID))
    for ID in phaseIDs
]

import numpy as np
import matplotlib.pyplot as plt
from pyrolite.util.plot.legend import proxy_line

np.random.seed(27)

fig, ax = plt.subplots(1)

for ix, (ID, style) in enumerate(zip(phaseIDs, styles)):
    ax.plot(np.arange(10), np.random.randn(10) + ix * 3, **style)

These are also handy for generating legend proxies, which can be used to generate summary legends where you may have multiple items with the same style:

proxies = [proxy_line(**sty) for sty in styles]

for ix, (ID, style) in enumerate(zip(phaseIDs, styles)):
    for i in range(3):  # make a few more lines per phaseID
        ax.plot(np.arange(10), np.random.randn(10) + ix * 3, **style)

ax.legend(  # use our proxy lines to generate a legend
    proxies,
    phaseIDs,
    frameon=False,
    facecolor=None,
    bbox_to_anchor=(1.0, 1.0),
    loc="upper left",
)
plt.show()