Ternary Plots

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd

from pyrolite.plot import pyroplot

np.random.seed(82)

Let’s first create some example data:

df = pd.DataFrame(data=np.exp(np.random.rand(100, 3)), columns=["SiO2", "MgO", "CaO"])
df.loc[:, ["SiO2", "MgO", "CaO"]].head()

	SiO2	MgO	CaO
0	1.316828	1.895023	1.866191
1	2.177206	1.613029	2.609435
2	2.711316	2.299090	1.226108
3	1.143450	1.849684	1.809385
4	1.905569	2.687793	1.398168

Now we can create a simple scatter plot:

ax = df.loc[:, ["SiO2", "MgO", "CaO"]].pyroplot.scatter(c="k")
plt.show()

If the data represent some continuous series, you could also plot them as lines:

ax = df.loc[:, ["SiO2", "MgO", "CaO"]].pyroplot.plot(color="k", alpha=0.5)
plt.show()

The plotting axis can be specified to use exisiting axes:

fig, ax = plt.subplots(1, 2, sharex=True, sharey=True, figsize=(12, 5))

df.loc[:, ["SiO2", "MgO", "CaO"]].sample(20).pyroplot.scatter(ax=ax[0], c="k")
df.loc[:, ["SiO2", "MgO", "CaO"]].sample(20).pyroplot.scatter(ax=ax[1], c="g")

ax = fig.orderedaxes  # creating scatter plots reorders axes, this is the correct order
plt.tight_layout()