Tiny little stupid showcase of how to plot multiple-group-timeseries as scatterplot, and click CLI library

timeseries_plot_using_matplotlib.py

# -*- coding: utf-8 -*-
"""visualizer: basic visualization module for debugging
"""

import pandas as pd
import matplotlib
import matplotlib.cm as cm
from matplotlib import pyplot as plt
from matplotlib import dates as mdates
import seaborn as sns
import click


def scatterplot(input, output):
    click.echo('Start processing of visualization')

    click.echo('Loading CSV file {}'.format(input))
    # Load N-Detection formatted CSV file
    df = pd.read_csv(input)

    # Convert date-string columns to datetime object
    df['TIME'] = pd.to_datetime(df['TIME'])

    # Setting plotting styles
    sns.set()
    sns.set_style("darkgrid")
    sns.set(font='Tahoma', font_scale=1.4)
    c = sns.color_palette("Set2", 10)

    f, ax = plt.subplots(figsize=(20, 10))
    for i, (n, g) in enumerate(df.groupby('GROUP')):
        plt.plot(g['TIME'], g['Y'], 'o', label=n, color=c[i % 10])

    ax.set_ylim(0, ax.get_ylim()[1] * 1.2)
    ax.set_ylabel('Y Axis label')
    ax.legend(loc="upper left", bbox_to_anchor=(1, 1))
    date_formatter = mdates.DateFormatter('%Y%m%d %H:%M')
    ax.xaxis.set_major_formatter(date_formatter)
    f.autofmt_xdate()

    # Output as PNG file
    f.savefig(output, bbox_inches='tight')

    click.echo('Write scatter-plot as {}'.format(output))
    return None


@click.command()
@click.option('--input', default=None, help='Path of CSV file to process')
@click.option('--output', default='scatterplot.png', help='Path to output file')
def cli(input, output):
    scatterplot(input, output)


if __name__ == '__main__':
    cli()