tabs/invalid_markets.py

import pandas as pd
import gradio as gr
import matplotlib.pyplot as plt
import seaborn as sns


def plot_daily_dist_invalid_trades(invalid_trades: pd.DataFrame):
    """Function to paint the distribution of daily invalid trades, no matter which market"""
    plot = sns.histplot(data=invalid_trades, x="creation_date", kde=True)
    plt.xticks(rotation=45, ha="right")
    plt.xlabel("Creation date")
    plt.ylabel("Daily number of invalid trades")
    plt.title("Distribution of daily invalid trades over time")
    return gr.Plot(value=plot.get_figure())


def plot_daily_nr_invalid_markets(invalid_trades: pd.DataFrame):
    """Function to paint the number of invalid markets over time"""
    daily_invalid_markets = (
        invalid_trades.groupby("creation_date")
        .agg(trades_count=("title", "count"), nr_markets=("title", "nunique"))
        .reset_index()
    )
    sns.set_theme(palette="viridis")
    plot = sns.lineplot(data=daily_invalid_markets, x="creation_date", y="nr_markets")
    plt.xticks(rotation=45, ha="right")
    plt.xlabel("Creation date")
    plt.ylabel("Daily number of invalid markets")
    plt.title("Evolution of daily invalid markets over time")
    return gr.Plot(value=plot.get_figure())


def plot_ratio_invalid_trades_per_market(invalid_trades: pd.DataFrame):
    """Function to paint the number of invalid trades that the same market accummulates"""
    cat = invalid_trades["title"]
    codes, uniques = pd.factorize(cat)

    # add the IDs as a new column to the original dataframe
    invalid_trades["title_id"] = codes
    plot = sns.displot(invalid_trades, x="title_id")
    plt.xlabel("market id")
    plt.ylabel("Total number of invalid trades by market")
    plt.title("Distribution of invalid trades per market")
    return gr.Plot(value=plot.get_figure())


def plot_top_invalid_markets(invalid_trades: pd.DataFrame):
    """Function to paint the top markets with the highest number of invalid trades"""
    top_invalid_markets = invalid_trades.title.value_counts().reset_index()
    top_invalid_markets.rename(columns={"count": "nr_invalid_trades"}, inplace=True)
    plt.figure(figsize=(25, 10))
    plot = sns.barplot(
        top_invalid_markets,
        x="nr_invalid_trades",
        y="title",
        hue="title",
        dodge=False,
    )
    return gr.Plot(value=plot.get_figure())