Plots distribution of scores for AVEC2017

plt_avec.py

import numpy as np
import pandas as pd
from fire import Fire
import plotnine as p


def plot(inputfile: str):
    """Plots using plotnine Test or dev destirbution of labels

    :inputfile: str: TODO
    :returns: TODO

    """
    df = pd.read_csv(inputfile)
    df['Healthy?'] = df['PHQ8_Binary'].apply(lambda x: "Yes" if x else "No")
    df['dummy'] = 1

    # desc = df['PHQ8_Score'].describe()
    print(p.ggplot(df, p.aes(x='dummy', y='PHQ8_Score', fill='Healthy?'))
          + p.geom_jitter(stat='identity', width=0.2,
                          alpha=0.9, show_legend=False, size=2)
          + p.ggtitle('Training data PHQ8 distribution')
          + p.labs(x="")
          + p.stat_summary(fun_y=np.mean, geom='point',
                           size=4, group='Healthy?', fun_ymin=np.min, fun_ymax=np.max)
          + p.coord_flip()
          # + stat_summary(aes(label='PHQ8_Score'), fun_y=np.mean, geom='text')
          + p.stat_summary(fun_data='mean_sdl',
                           fun_args={'mult': 1}, geom='errorbar', size=0.2)
          + p.scale_x_continuous(labels=[], breaks=[])
          + p.theme(text=p.element_text(family='Droid Sans Fallback'),
                    figure_size=(10, 4))
          )


if __name__ == "__main__":
    Fire(plot)