imas_timeline.py

import colorsys
import csv
from datetime import datetime
from pathlib import Path

import matplotlib.colors as mcolors
import matplotlib.pyplot as plt
import numpy as np
from matplotlib import markers
from matplotlib.path import Path as pPath


# https://stackoverflow.com/questions/26686722/align-matplotlib-scatter-marker-left-and-or-right
def align_marker(marker, halign='center', valign='middle'):
    if isinstance(halign, (str)):
        halign = {'right': -1.,
                  'middle': 0.,
                  'center': 0.,
                  'left': 1.,
                  }[halign]

    if isinstance(valign, (str)):
        valign = {'top': -1.,
                  'middle': 0.,
                  'center': 0.,
                  'bottom': 1.,
                  }[valign]

    # Define the base marker
    bm = markers.MarkerStyle(marker)

    # Get the marker path and apply the marker transform to get the
    # actual marker vertices (they should all be in a unit-square
    # centered at (0, 0))
    m_arr = bm.get_path().transformed(bm.get_transform()).vertices

    # Shift the marker vertices for the specified alignment.
    m_arr[:, 0] += halign / 2
    m_arr[:, 1] += valign / 2

    return pPath(m_arr, bm.get_path().codes)

def tsv2dict(p):
    p = Path(p)
    with p.open(encoding='utf8') as f:
        reader = csv.DictReader(f, delimiter='\t', quoting=csv.QUOTE_NONE)
        # headers = reader.fieldnames
        data = list(reader)
    return data

def adjust_lightness(color, amount=0.5):
    try:
        c = mcolors.cnames[color]
    except:
        c = color
    c = colorsys.rgb_to_hls(*mcolors.to_rgb(c))
    return colorsys.hls_to_rgb(c[0], max(0, min(1, amount * c[1])), c[2])

def ordinal(n: int):
    if 11 <= (n % 100) <= 13:
        suffix = 'th'
    else:
        suffix = ['th', 'st', 'nd', 'rd', 'th'][min(n % 10, 4)]
    return str(n) + suffix

def cleanup():
    '''
    Clean up the data copied from https://music765plus.com/%E3%83%A9%E3%82%A4%E3%83%96%26%E6%AD%8C%E3%81%82%E3%82%8A%E3%82%A4%E3%83%99%E3%83%B3%E3%83%88%E3%81%AE%E4%B8%80%E8%A6%A7
    I still have to manually clean more later in the live_event_data_new.tsv and save the result to live_event_data_cleaned.tsv
    '''
    data = Path('live_event_data.tsv').read_text(encoding='utf-8').splitlines()
    data_new = []

    for d in data:
        cat, date, title = d.split('\t')
        date = date.lstrip('D')
        for no in range(1, 11):
            if ordinal(no) in title.lower():
                print(d, 'has ordinal', ordinal(no))
                title += '\t' + ordinal(no)
        data_new.append('\t'.join([cat, date, title]))
    Path('live_event_data_new.tsv').write_text('\n'.join(data_new), encoding='utf-8')

def plot(by_days=False):
    data = tsv2dict('live_event_data_cleaned.tsv')
    # preprocess data
    data_ = []
    for d in data:
        date = d['date']
        if '~' in date:
            date = date.split('~')[0]
        date = datetime.strptime(date, '%Y-%m-%d')
        prev_data = data_[-1] if data_ else None
        if prev_data and prev_data['date'].day + 1 == date.day and d['number'] == prev_data['number']:
            # combine two-day events datapoint into one
            print('Second day of the same event! Ignore:')
            print('  ', d['event_name'])
            print('  ', prev_data['event_name'])
        else:
            d['date'] = date
            data_.append(d)
    data = data_

    CATS = ("[765]", "[CIN]", "[MIL]", "[SdM]", "[SnC]")
    PROPER_NAMES = ("765AS", "CINDERELLA GIRLS", "MILLION LIVE!", "SideM", "SHINY COLORS")
    COLORS = ('tab:red', 'tab:blue', 'tab:orange', 'tab:green', 'tab:cyan')

    # get the largest interval for x-axis
    all_dates = [d['date'] for d in data]
    if by_days:
        intervals = []
        for cat in CATS:
            data_ = [d for d in data if d['cat'] == cat]
            dates_ = [d['date'] for d in data_]
            interval = (dates_[-1] - dates_[0]).days
            intervals.append(interval)
        largest_intervals = max(intervals)
        x_base, y_base = 0, largest_intervals
    else:
        x_base, y_base = all_dates[0], all_dates[-1]

    fig, axes = plt.subplots(5, 1, figsize=(10, 5), layout="constrained")
    for idx, cat in enumerate(CATS):
        ax = axes[idx]
        ax.set(title=PROPER_NAMES[idx])

        data_ = [d for d in data if d['cat'] == cat]
        dates_ = [d['date'] for d in data_]
        if by_days:
            day0_ = dates_[0]
            for d in data_:
                d['date'] = (d['date'] - day0_).days

        # draw a baseline
        ax.hlines(0, data_[0]['date'], data_[-1]['date'], color=COLORS[idx], zorder=1)

        # split the data into two groups by alternating the event names
        data_groups =[[], []]
        prev = None
        current_group = 1
        for d in data_:
            number = d["number"]
            if number == prev:
                pass
            else:
                ax.annotate(number, xy=(d['date'], 0.5), xytext=(0, 0), textcoords="offset points", horizontalalignment="center", verticalalignment="bottom")
                current_group = 1 - current_group
            data_groups[current_group].append(d)
            prev = number

        for j, data__ in enumerate(data_groups):
            if j == 0:
                color = COLORS[idx]
            else:
                color = adjust_lightness(COLORS[idx], 1.5)
            dates__ = [d['date'] for d in data__]
            ax.scatter(dates__, np.zeros_like(dates__), marker=align_marker('v', valign='bottom'), s=600, color=color, edgecolors='black', zorder=3, clip_on=False)
            prev = None

        ax.set_xlim(x_base, y_base)
        ax.set_ylim(-1, 1)
        ax.yaxis.set_visible(False)

        # hide all lines but the x-axis for the bottom subplot
        if idx < 4:
            ax.xaxis.set_visible(False)
            ax.spines[["left", "top", "right", "bottom"]].set_visible(False)
        else:
            ax.spines[["left", "top", "right"]].set_visible(False)

    # plt.show()
    if by_days:
        filename = 'live_event_timeline_by_days.png'
    else:
        filename = 'live_event_timeline_by_date.png'

    plt.savefig(filename, dpi=150, bbox_inches='tight')

# cleanup()
plot(by_days=False)
plot(by_days=True)

live_event_data_cleaned.tsv