Added explanation text for divergence

app.py

@@ -22,8 +22,6 @@ from tabs.trader_plots import (
 
 from tabs.market_plots import (
     plot_kl_div_per_market,
-    plot_kl_div_with_off_by,
-    plot_kl_div_per_market2,
 )
 
 
@@ -203,7 +201,10 @@ with demo:
             with gr.Row():
                 trade_details_text = get_metrics_text()
             with gr.Row():
-                kl_div_plot = plot_kl_div_per_market2(closed_markets=closed_markets)
+                with gr.Column(scale=3):
+                    kl_div_plot = plot_kl_div_per_market(closed_markets=closed_markets)
+                with gr.Column(scale=1):
+                    interpretation = get_interpretation_text()
 
         with gr.TabItem("🎖️Weekly winning trades % per trader"):
             with gr.Row():
@@ -211,6 +212,7 @@ with demo:
             with gr.Row():
                 metrics_text = get_metrics_text()
             with gr.Row():
+
                 winning_metric = plot_winning_metric_per_trader(weekly_winning_metrics)
 
 demo.queue(default_concurrency_limit=40).launch()

notebooks/closed_markets.ipynb

@@ -1899,20 +1899,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 48,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "4.605170185988092\n"
+      "0.5108256237659907\n"
      ]
     }
    ],
    "source": [
     "P = np.array([1.0, 0.0])\n",
-    "Q = np.array([0.01, 0.99])\n",
+    "Q = np.array([0.60, 0.05])\n",
     "print(kl_divergence(P,Q))"
    ]
   },

notebooks/divergence.ipynb

@@ -0,0 +1,321 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "/Users/cyberosa/.pyenv/versions/hf_dashboards/lib/python3.12/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
+      "  from .autonotebook import tqdm as notebook_tqdm\n"
+     ]
+    }
+   ],
+   "source": [
+    "import pandas as pd\n",
+    "import gradio as gr\n",
+    "import plotly.express as px\n",
+    "import plotly.graph_objects as go\n",
+    "from plotly.subplots import make_subplots\n",
+    "import matplotlib.pyplot as plt\n",
+    "import seaborn as sns"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "div_data = pd.read_parquet(\"../data/closed_markets_div.parquet\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>currentAnswer</th>\n",
+       "      <th>id</th>\n",
+       "      <th>openingTimestamp</th>\n",
+       "      <th>market_creator</th>\n",
+       "      <th>opening_datetime</th>\n",
+       "      <th>first_outcome_prob</th>\n",
+       "      <th>second_outcome_prob</th>\n",
+       "      <th>kl_divergence</th>\n",
+       "      <th>off_by_perc</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>315</th>\n",
+       "      <td>no</td>\n",
+       "      <td>0x29462bf8c8f24772cd6da03878a4aee5c5813474</td>\n",
+       "      <td>1724976000</td>\n",
+       "      <td>pearl</td>\n",
+       "      <td>2024-08-30 02:00:00</td>\n",
+       "      <td>0.9416</td>\n",
+       "      <td>0.0584</td>\n",
+       "      <td>2.840439</td>\n",
+       "      <td>94.16</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>323</th>\n",
+       "      <td>yes</td>\n",
+       "      <td>0x0ad9d4edb0a401ec9a5b4f2ccf7942d28c29d4e3</td>\n",
+       "      <td>1724976000</td>\n",
+       "      <td>quickstart</td>\n",
+       "      <td>2024-08-30 02:00:00</td>\n",
+       "      <td>0.0499</td>\n",
+       "      <td>0.9501</td>\n",
+       "      <td>2.997734</td>\n",
+       "      <td>95.01</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    currentAnswer                                          id  \\\n",
+       "315            no  0x29462bf8c8f24772cd6da03878a4aee5c5813474   \n",
+       "323           yes  0x0ad9d4edb0a401ec9a5b4f2ccf7942d28c29d4e3   \n",
+       "\n",
+       "    openingTimestamp market_creator    opening_datetime  first_outcome_prob  \\\n",
+       "315       1724976000          pearl 2024-08-30 02:00:00              0.9416   \n",
+       "323       1724976000     quickstart 2024-08-30 02:00:00              0.0499   \n",
+       "\n",
+       "     second_outcome_prob  kl_divergence  off_by_perc  \n",
+       "315               0.0584       2.840439        94.16  \n",
+       "323               0.9501       2.997734        95.01  "
+      ]
+     },
+     "execution_count": 3,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "div_data.loc[div_data[\"off_by_perc\"]>=90]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/html": [
+       "<div>\n",
+       "<style scoped>\n",
+       "    .dataframe tbody tr th:only-of-type {\n",
+       "        vertical-align: middle;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe tbody tr th {\n",
+       "        vertical-align: top;\n",
+       "    }\n",
+       "\n",
+       "    .dataframe thead th {\n",
+       "        text-align: right;\n",
+       "    }\n",
+       "</style>\n",
+       "<table border=\"1\" class=\"dataframe\">\n",
+       "  <thead>\n",
+       "    <tr style=\"text-align: right;\">\n",
+       "      <th></th>\n",
+       "      <th>currentAnswer</th>\n",
+       "      <th>id</th>\n",
+       "      <th>openingTimestamp</th>\n",
+       "      <th>market_creator</th>\n",
+       "      <th>opening_datetime</th>\n",
+       "      <th>first_outcome_prob</th>\n",
+       "      <th>second_outcome_prob</th>\n",
+       "      <th>kl_divergence</th>\n",
+       "      <th>off_by_perc</th>\n",
+       "    </tr>\n",
+       "  </thead>\n",
+       "  <tbody>\n",
+       "    <tr>\n",
+       "      <th>52</th>\n",
+       "      <td>no</td>\n",
+       "      <td>0x927beda324bfd4514a7b64ab5594451fdaf4796e</td>\n",
+       "      <td>1722816000</td>\n",
+       "      <td>quickstart</td>\n",
+       "      <td>2024-08-05 02:00:00</td>\n",
+       "      <td>0.8792</td>\n",
+       "      <td>0.1208</td>\n",
+       "      <td>2.113619</td>\n",
+       "      <td>87.92</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>293</th>\n",
+       "      <td>yes</td>\n",
+       "      <td>0x90bb15982f2b5a5f044ad8ff49fe20daddfb8ca7</td>\n",
+       "      <td>1724803200</td>\n",
+       "      <td>quickstart</td>\n",
+       "      <td>2024-08-28 02:00:00</td>\n",
+       "      <td>0.1166</td>\n",
+       "      <td>0.8834</td>\n",
+       "      <td>2.149006</td>\n",
+       "      <td>88.34</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>315</th>\n",
+       "      <td>no</td>\n",
+       "      <td>0x29462bf8c8f24772cd6da03878a4aee5c5813474</td>\n",
+       "      <td>1724976000</td>\n",
+       "      <td>pearl</td>\n",
+       "      <td>2024-08-30 02:00:00</td>\n",
+       "      <td>0.9416</td>\n",
+       "      <td>0.0584</td>\n",
+       "      <td>2.840439</td>\n",
+       "      <td>94.16</td>\n",
+       "    </tr>\n",
+       "    <tr>\n",
+       "      <th>323</th>\n",
+       "      <td>yes</td>\n",
+       "      <td>0x0ad9d4edb0a401ec9a5b4f2ccf7942d28c29d4e3</td>\n",
+       "      <td>1724976000</td>\n",
+       "      <td>quickstart</td>\n",
+       "      <td>2024-08-30 02:00:00</td>\n",
+       "      <td>0.0499</td>\n",
+       "      <td>0.9501</td>\n",
+       "      <td>2.997734</td>\n",
+       "      <td>95.01</td>\n",
+       "    </tr>\n",
+       "  </tbody>\n",
+       "</table>\n",
+       "</div>"
+      ],
+      "text/plain": [
+       "    currentAnswer                                          id  \\\n",
+       "52             no  0x927beda324bfd4514a7b64ab5594451fdaf4796e   \n",
+       "293           yes  0x90bb15982f2b5a5f044ad8ff49fe20daddfb8ca7   \n",
+       "315            no  0x29462bf8c8f24772cd6da03878a4aee5c5813474   \n",
+       "323           yes  0x0ad9d4edb0a401ec9a5b4f2ccf7942d28c29d4e3   \n",
+       "\n",
+       "    openingTimestamp market_creator    opening_datetime  first_outcome_prob  \\\n",
+       "52        1722816000     quickstart 2024-08-05 02:00:00              0.8792   \n",
+       "293       1724803200     quickstart 2024-08-28 02:00:00              0.1166   \n",
+       "315       1724976000          pearl 2024-08-30 02:00:00              0.9416   \n",
+       "323       1724976000     quickstart 2024-08-30 02:00:00              0.0499   \n",
+       "\n",
+       "     second_outcome_prob  kl_divergence  off_by_perc  \n",
+       "52                0.1208       2.113619        87.92  \n",
+       "293               0.8834       2.149006        88.34  \n",
+       "315               0.0584       2.840439        94.16  \n",
+       "323               0.9501       2.997734        95.01  "
+      ]
+     },
+     "execution_count": 5,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "div_data.loc[div_data[\"kl_divergence\"]>=2.0]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "all_markets = closed_markets.copy(deep=True)\n",
+    "    all_markets[\"market_creator\"] = \"all\"\n",
+    "\n",
+    "    # merging both dataframes\n",
+    "    final_markets = pd.concat([div_data, all_markets], ignore_index=True)\n",
+    "    final_markets = final_markets.sort_values(by=\"opening_datetime\", ascending=True)\n",
+    "\n",
+    "    # Create the main figure and axis\n",
+    "    fig, ax1 = plt.subplots(figsize=(10, 6))\n",
+    "\n",
+    "    # Create the boxplot using seaborn\n",
+    "    sns.boxplot(\n",
+    "        data=closed_markets,\n",
+    "        x=\"month_year_week\",\n",
+    "        y=\"kl_divergence\",\n",
+    "        ax=ax1,\n",
+    "        hue=\"market_creator\",\n",
+    "        order=[\"pearl\", \"quickstart\", \"all\"],\n",
+    "    )\n",
+    "\n",
+    "    # Set labels and title for the main axis\n",
+    "    ax1.set_xlabel(\"Week\")\n",
+    "    ax1.set_ylabel(\"KL Divergence\")\n",
+    "    ax1.set_title(\"KL Divergence Boxplot with Off-by Percentage\")\n",
+    "\n",
+    "    # Create a secondary y-axis\n",
+    "    ax2 = ax1.twinx()\n",
+    "\n",
+    "    # Plot the off_by_perc values on the secondary y-axis\n",
+    "    for i, week in enumerate(closed_markets[\"month_year_week\"].unique()):\n",
+    "        off_by_perc = closed_markets[closed_markets[\"month_year_week\"] == week][\n",
+    "            \"off_by_perc\"\n",
+    "        ]\n",
+    "        ax2.scatter([i] * len(off_by_perc), off_by_perc, color=\"red\", alpha=0.01)\n",
+    "\n",
+    "    # Set label for the secondary y-axis\n",
+    "    ax2.set_ylabel(\"Off-by Percentage\")\n",
+    "\n",
+    "    # Adjust the layout and display the plot\n",
+    "    plt.tight_layout()"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "hf_dashboards",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.12.2"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

tabs/market_plots.py

@@ -30,6 +30,8 @@ def plot_kl_div_per_market(closed_markets: pd.DataFrame) -> gr.Plot:
         xaxis_title="Markets closing Week",
         yaxis_title="Kullback–Leibler divergence",
         legend=dict(yanchor="top", y=0.5),
+        width=1000,  # Adjusted for better fit on laptop screens
+        height=600,  # Adjusted for better fit on laptop screens
     )
 
     fig.update_xaxes(tickformat="%b %d\n%Y")
@@ -83,77 +85,3 @@ def plot_kl_div_with_off_by(closed_markets: pd.DataFrame) -> gr.Plot:
     return gr.Plot(
         value=fig,
     )
-
-
-def plot_kl_div_per_market2(closed_markets: pd.DataFrame) -> gr.Plot:
-
-    # adding the total
-    all_markets = closed_markets.copy(deep=True)
-    all_markets["market_creator"] = "all"
-
-    # merging both dataframes
-    final_markets = pd.concat([closed_markets, all_markets], ignore_index=True)
-    final_markets = final_markets.sort_values(by="opening_datetime", ascending=True)
-
-    # Create the figure with secondary y-axis
-    fig = make_subplots(specs=[[{"secondary_y": True}]])
-
-    # Add box plots for KL divergence
-    for creator in ["pearl", "quickstart", "all"]:
-        fig.add_trace(
-            go.Box(
-                x=final_markets[final_markets["market_creator"] == creator][
-                    "month_year_week"
-                ],
-                y=final_markets[final_markets["market_creator"] == creator][
-                    "kl_divergence"
-                ],
-                name=creator,
-                boxmean=True,
-                marker_color={
-                    "pearl": "purple",
-                    "quickstart": "goldenrod",
-                    "all": "darkgreen",
-                }[creator],
-            ),
-            secondary_y=False,
-        )
-        fig.add_trace(
-            go.Scatter(
-                x=final_markets[final_markets["market_creator"] == creator][
-                    "month_year_week"
-                ],
-                y=final_markets[final_markets["market_creator"] == creator][
-                    "off_by_perc"
-                ],
-                opacity=0.01,
-                showlegend=False,
-                hoverinfo="skip",
-            ),
-            secondary_y=True,
-        )
-    # Update layout
-    fig.update_layout(
-        xaxis_title="Markets closing Week",
-        # legend=dict(yanchor="top", y=0.5, xanchor="left", x=-0.5),
-        legend=dict(yanchor="top", y=0.5),
-        boxmode="group",
-        width=1000,  # Adjusted for better fit on laptop screens
-        height=600,  # Adjusted for better fit on laptop screens
-    )
-    # Update y-axes
-    fig.update_yaxes(
-        title_text="Kullback–Leibler divergence", secondary_y=False, side="left"
-    )
-    fig.update_yaxes(
-        title_text="Off by percentage between probabilities",
-        secondary_y=True,
-        side="right",
-    )
-
-    # Update x-axis
-    fig.update_xaxes(tickformat="%b %d\n%Y")
-
-    return gr.Plot(
-        value=fig,
-    )

tabs/trader_plots.py

@@ -26,28 +26,19 @@ def get_metrics_text() -> gr.Markdown:
 
 def get_interpretation_text() -> gr.Markdown:
     interpretation_text = """
-        ## Interpreting KL-divergence values
-            ### Moderate divergence:
-            * Y ≈ 0.9163
-                * Wrong market prediction closing at [0.4, 0.6] or [0.6, 0.4]
-            * Y ≈ 1.2040
-                * Wrong market prediction closing at [0.3, 0.7] or [0.7, 0.3]
-            * Y ≈ 1.6093
-                * Wrong market prediction closing at [0.2, 0.8] or [0.8, 0.2]
-            * Y ≈ 2.3026
-                * Wrong market prediction closing at [0.1, 0.9] or [0.9, 0.1]
-
-            ### High divergence:
-            * Y ≈ 4.6052
-                * Very wrong prediction closing at [0.01, 0.99] or [0.99, 0.01]
-            * Y ≈ 6.9078
-                * Extremely wrong prediction closing at [0.001, 0.999] or [0.999, 0.001]
-
-            ### Edge cases:
-            * Y ≈ 13.8155
-                * Near-certain wrong prediction closing at [1e-6, 0.999999] or [0.999999, 1e-6]
-            * Y = 20 (capped)
-                * Represents cases where the market prediction was essentially opposite to the outcome.
+        ## Meaning of KL-divergence values
+            * Y = 0.05129
+                * Market accuracy off by 5%
+            * Y = 0.1053
+                * Market accuracy off by 10%
+            * Y = 0.2876
+                * Market accuracy off by 25%
+            * Y = 0.5108
+                * Market accuracy off by 40%
+            * Y = 1.2040
+                * Market accuracy off by 70%
+            * Y = 2.3026
+                * Market accuracy off by 90%
     """
     return gr.Markdown(interpretation_text)