Turn ha and va into aesthetics for geom_text()

[jeroenjanssens]

While recreating the following figure of Section 28.3 from R for Data Science:

I noticed that ha and va are not aesthetics but parameters. According to the documentation, this is because of a limitation in Matplotlib (see matplotlib/matplotlib#1181). Unless I fail to fully understand the issue, I believe that this small change fixes it. At least I was able to reproduce the above figure with the following code:

from plotnine import *
import pandas as pd
from itertools import product

has = ["left", "center", "right"]
vas = ["top", "center", "bottom"]
xs = [0, 0.5, 1]
ys = [1, 0.5, 0]

df = pd.DataFrame([{"x": xs[x],
                    "y": ys[y],
                    "ha": has[x],
                    "va": vas[y],
                    "label": f"ha=\"{has[x]}\"\nva=\"{vas[y]}\""}
                   for x, y in product(range(3), repeat=2)])

ggplot(df, aes("x", "y")) +\
geom_point(colour="grey", size=5) +\
geom_point(size=0.5, colour="red") +\
geom_text(aes(label="label", ha="ha", va="va"), size=14) +\
labs(x=None, y=None)

Please forgive me for not yet taking care of the deprecated terms "hjust" and "vjust". I wasn't able to quickly figure out how to treat those as aliases in the case of aesthetics.

[has2k1]

This PR also raises the issue that the aesthetics are not documented (with descriptions). So far, I think all the aesthetics have been self-descriptive, but va and ha are not.

[has2k1]

This PR also raises the issue that the aesthetics are not documented (with descriptions). So far, I think all the aesthetics have been self-descriptive, but va and ha are not.

I was wrong. There is a provision for documenting them as some stats do. I had forgotten about it!

[has2k1]

@jeroenjanssens. Thanks. I will handle the documentation and the old parameters names.