Linear regression line on facet grid with logarithmic Y axis

Question

I have code working exactly as I want: it plots two colored datasets, TypeA and TypeB, on three plots separated by the value Divison with a logarithmic axis for Y only.

pal = dict(TypeA="seagreen", TypeB="gray")
g = sns.FacetGrid(df, hue="Source", col='Division', palette=pal, size=5)
g.map(plt.scatter, "X_Data", "Y_LogData")
g.add_legend();
g.set(yscale="log")

How can I get a linear regression line to plot?

Answer 1

Matplotlib's plt.scatter does not calculate a regression line. Pass sns.regplot to g.map() instead, which calculates a regression line by default.

You can see some examples here where different plotting functions are used with FacetGrid.map() (which is g.map() in your code).