tabs/metrics.py

import pandas as pd
import gradio as gr
import plotly.express as px
import gc
import matplotlib.pyplot as plt

trade_metric_choices = [
    "mech calls",
    "collateral amount",
    "earnings",
    "net earnings",
    "ROI",
]

tool_metric_choices = {
    "Weekly Mean Mech Tool Accuracy as (Accurate Responses/All) %": "win_perc",
    "Total Weekly Inaccurate Nr of Mech Tool Responses": "losses",
    "Total Weekly Accurate Nr of Mech Tool Responses": "wins",
    "Total Weekly Nr of Mech Tool Requests": "total_request",
}

default_trade_metric = "ROI"
default_tool_metric = "Weekly Mean Mech Tool Accuracy as (Accurate Responses/All) %"

HEIGHT = 600
WIDTH = 1000


def get_metrics(
    metric_name: str, column_name: str, market_creator: str, trades_df: pd.DataFrame
) -> pd.DataFrame:
    # this is to filter out the data before 2023-09-01
    trades_filtered = trades_df[trades_df["creation_timestamp"] > "2023-09-01"]
    if market_creator != "all":
        trades_filtered = trades_filtered.loc[
            trades_filtered["market_creator"] == market_creator
        ]

    trades_filtered = (
        trades_filtered.groupby("month_year_week", sort=False)[column_name]
        .quantile([0.25, 0.5, 0.75])
        .unstack()
    )
    # reformat the data as percentile, date, value
    trades_filtered = trades_filtered.melt(
        id_vars=["month_year_week"], var_name="percentile", value_name=metric_name
    )
    trades_filtered.columns = trades_filtered.columns.astype(str)
    trades_filtered.reset_index(inplace=True)
    trades_filtered.columns = [
        "month_year_week",
        "25th_percentile",
        "50th_percentile",
        "75th_percentile",
    ]
    # reformat the data as percentile, date, value
    trades_filtered = trades_filtered.melt(
        id_vars=["month_year_week"], var_name="percentile", value_name=metric_name
    )
    return trades_filtered


def get_boxplot_metrics(column_name: str, trades_df: pd.DataFrame) -> pd.DataFrame:
    trades_filtered = trades_df[
        ["creation_timestamp", "month_year_week", "market_creator", column_name]
    ]
    # adding the total
    trades_filtered_all = trades_df.copy(deep=True)
    trades_filtered_all["market_creator"] = "all"

    # merging both dataframes
    all_filtered_trades = pd.concat(
        [trades_filtered, trades_filtered_all], ignore_index=True
    )
    all_filtered_trades = all_filtered_trades.sort_values(
        by="creation_timestamp", ascending=True
    )
    gc.collect()
    return all_filtered_trades


def plot2_trade_details(
    metric_name: str, market_creator: str, trades_df: pd.DataFrame
) -> gr.Plot:
    """Plots the trade details for the given trade detail."""

    if metric_name == "mech calls":
        metric_name = "mech_calls"
        column_name = "num_mech_calls"
        yaxis_title = "Nr of mech calls per trade"
    elif metric_name == "ROI":
        column_name = "roi"
        yaxis_title = "ROI (net profit/cost)"
    elif metric_name == "collateral amount":
        metric_name = "collateral_amount"
        column_name = metric_name
        yaxis_title = "Collateral amount per trade (xDAI)"
    elif metric_name == "net earnings":
        metric_name = "net_earnings"
        column_name = metric_name
        yaxis_title = "Net profit per trade (xDAI)"
    else:  # earnings
        column_name = metric_name
        yaxis_title = "Gross profit per trade (xDAI)"

    trades_filtered = get_metrics(metric_name, column_name, market_creator, trades_df)
    fig = px.line(
        trades_filtered, x="month_year_week", y=metric_name, color="percentile"
    )
    fig.update_layout(
        xaxis_title="Week",
        yaxis_title=yaxis_title,
        legend=dict(yanchor="top", y=0.5),
    )
    fig.update_xaxes(tickformat="%b %d\n%Y")
    return gr.Plot(
        value=fig,
    )


def plot_trade_metrics(metric_name: str, trades_df: pd.DataFrame) -> gr.Plot:
    """Plots the trade metrics."""

    if metric_name == "mech calls":
        metric_name = "mech_calls"
        column_name = "num_mech_calls"
        yaxis_title = "Nr of mech calls per trade"
    elif metric_name == "ROI":
        column_name = "roi"
        yaxis_title = "ROI (net profit/cost)"
    elif metric_name == "collateral amount":
        metric_name = "collateral_amount"
        column_name = metric_name
        yaxis_title = "Collateral amount per trade (xDAI)"
    elif metric_name == "net earnings":
        metric_name = "net_earnings"
        column_name = metric_name
        yaxis_title = "Net profit per trade (xDAI)"
    else:  # earnings
        column_name = metric_name
        yaxis_title = "Gross profit per trade (xDAI)"

    trades_filtered = get_boxplot_metrics(column_name, trades_df)
    fig = px.box(
        trades_filtered,
        x="month_year_week",
        y=column_name,
        color="market_creator",
        color_discrete_sequence=["purple", "goldenrod", "darkgreen"],
        category_orders={"market_creator": ["pearl", "quickstart", "all"]},
    )
    fig.update_traces(boxmean=True)
    fig.update_layout(
        xaxis_title="Week",
        yaxis_title=yaxis_title,
        legend=dict(yanchor="top", y=0.5),
    )
    fig.update_xaxes(tickformat="%b %d\n%Y")
    return gr.Plot(
        value=fig,
    )


def plot_average_roi_per_market_by_week(trades_df: pd.DataFrame) -> gr.LinePlot:

    mean_roi_per_market_by_week = (
        trades_df.groupby(["market_creator", "month_year_week"])["roi"]
        .mean()
        .reset_index()
    )
    mean_roi_per_market_by_week.rename(columns={"roi": "mean_roi"}, inplace=True)
    return gr.LinePlot(
        value=mean_roi_per_market_by_week,
        x="month_year_week",
        y="ROI",
        color="market_creator",
        show_label=True,
        interactive=True,
        show_actions_button=True,
        tooltip=["month_year_week", "market_creator", "mean_roi"],
        height=HEIGHT,
        width=WIDTH,
    )


def get_trade_metrics_text() -> gr.Markdown:
    metric_text = """ 
        ## Description of the graph
        These metrics are computed weekly. The statistical measures are:
        * min, max, 25th(q1), 50th(median) and 75th(q2) percentiles
        * the upper and lower fences to delimit possible outliers
        * the average values as the dotted lines
        """

    return gr.Markdown(metric_text)