plotter.py

from lxml import html

string = ""

# https://www.oddschecker.com/politics/us-politics/us-presidential-election-2020/winner/bet-history/bernie-sanders#all-history
# with open("table-bernie.html", "r") as f:
#     string = f.read()

# https://www.oddschecker.com/politics/us-politics/us-presidential-election-2020/winner/bet-history/joe-biden#all-history
# with open("table-biden.html", "r") as f:
#     string = f.read()

# https://www.oddschecker.com/politics/us-politics/us-presidential-election-2020/winner/bet-history/elizabeth-warren#all-history
# with open("table-warren.html", "r") as f:
#     string = f.read()

# https://www.oddschecker.com/politics/us-politics/us-presidential-election-2020/winner/bet-history/donald-trump#all-history

LIST_OF_NAMES = ["bernie-sanders", "pete-buttigieg", "elizabeth-warren", "joe-biden", "donald-trump"]
name_to_data = {}
name_to_inverse_data = {}
for name in LIST_OF_NAMES:
    with open("./5-feb/%s" % name, "r") as f:
        string = f.read()

    tree = html.fromstring(string)

    num_rows = len(tree.xpath("/html/body/div[1]/div/div/div/div/div/section/div/div/div/div[3]/div[2]/table/tbody/*"))

    all_days_of_betfair_odds = []
    all_date_strings = []
    for i in range(1, num_rows):
        # 22 is hardcoded for Betfair's column
        one_day_of_betfair_odds = tree.xpath("/html/body/div[1]/div/div/div/div/div/section/div/div/div/div[3]/div[2]/table/tbody/tr[%d]/td[22]/*/text()" % i)
        all_days_of_betfair_odds.append(one_day_of_betfair_odds)
        date_string = tree.xpath("/html/body/div[1]/div/div/div/div/div/section/div/div/div/div[3]/div[2]/table/tbody/tr[%d]/*/text()" % i)[0]
        all_date_strings.append(date_string)

    assert len(all_days_of_betfair_odds) == len(all_date_strings)

    all_days_of_betfair_odds_decimals = []
    for one_day_of_betfair_odds in all_days_of_betfair_odds:
        one_day_of_betfair_odds_decimals = []
        for one_odds in one_day_of_betfair_odds:
            try:
                numerator, denominator = one_odds.split("/")
            except ValueError:
                numerator = int(one_odds)
                denominator = 1
            decimal = int(numerator) / int(denominator)
            one_day_of_betfair_odds_decimals.append(decimal)
        all_days_of_betfair_odds_decimals.append(one_day_of_betfair_odds_decimals)

    all_date_strings_with_data = []
    all_opens = []
    all_highs = []
    all_lows = []
    all_closes = []
    for i in range(0, len(all_days_of_betfair_odds_decimals)):
        try:
            one_date_string_with_data = all_date_strings[i]
            one_open = all_days_of_betfair_odds_decimals[i][-1]
            one_high = max(all_days_of_betfair_odds_decimals[i])
            one_low = min(all_days_of_betfair_odds_decimals[i])
            one_close = all_days_of_betfair_odds_decimals[i][0]

            all_date_strings_with_data.append(one_date_string_with_data)
            all_opens.append(one_open)
            all_highs.append(one_high)
            all_lows.append(one_low)
            all_closes.append(one_close)
        except IndexError:
            print("incomplete data for %s" % all_date_strings[i])

    assert len(all_date_strings_with_data) == \
            len(all_opens) == len(all_highs) == len(all_closes)

    import plotly.graph_objects as go
    import pandas as pd
    from datetime import datetime

    # XXX is this right? 1/1 odds -> 50/50% yeah?
    all_inverse_opens = list(map(lambda x: 1.0 / (1 + x), all_opens))
    all_inverse_highs = list(map(lambda x: 1.0 / (1 + x), all_highs))
    all_inverse_lows = list(map(lambda x: 1.0 / (1 + x), all_lows))
    all_inverse_closes = list(map(lambda x: 1.0 / (1 + x), all_closes))

    # fig1 = go.Figure(
    #     data=[
    #         go.Candlestick(
    #             x=all_date_strings_with_data,
    #             open=all_inverse_opens,
    #             high=all_inverse_highs,
    #             low=all_inverse_lows,
    #             close=all_inverse_closes,
    #         ),
    #     ]
    # )
    # fig1.update_layout(title=name)
    # fig1.update_layout(
    #         xaxis=go.layout.XAxis(
    #             fixedrange=True,
    #             range=["2019-10-01 00:00:00.0000", "2020-02-06 00:00:00.0000"],
    #             rangeslider=dict(visible=False),
    #             type="date"
    #         ),
    #         yaxis=go.layout.YAxis(
    #             autorange=True
    #         )
    # )
    # fig1.show()

    name_to_inverse_data[name] = {}
    name_to_inverse_data[name]["dates"] = all_date_strings_with_data
    name_to_inverse_data[name]["opens"] = all_inverse_opens
    name_to_inverse_data[name]["highs"] = all_inverse_highs
    name_to_inverse_data[name]["lows"] = all_inverse_lows
    name_to_inverse_data[name]["closes"] = all_inverse_closes


import pandas as pd
from plotly.subplots import make_subplots
from datetime import datetime
import plotly.express as px


# fig = make_subplots(
#             rows=len(LIST_OF_NAMES), cols=1,
#             subplot_titles=tuple(LIST_OF_NAMES),
#             shared_xaxes=True,
#             vertical_spacing=0.1)

fig = go.Figure()

row_num = 0
for name in LIST_OF_NAMES:
    color1 = px.colors.qualitative.Dark2[row_num]
    # color2 = px.colors.qualitative.Set2[row_num]
    fig.add_trace(
        go.Candlestick(x=name_to_inverse_data[name]['dates'],
                open=name_to_inverse_data[name]["opens"],
                high=name_to_inverse_data[name]["highs"],
                low=name_to_inverse_data[name]["lows"],
                close=name_to_inverse_data[name]["closes"],
                increasing_line_color=color1,
                decreasing_line_color=color1,
                name=name,
                ),
    )
    row_num = row_num + 1

fig.update_layout(
        xaxis=go.layout.XAxis(
            fixedrange=True,
            range=["2019-09-01 00:00:00.0000", "2020-02-06 00:00:00.0000"],
            rangeslider=dict(visible=False),
            type="date"
        )
)
fig.show()