Merge pull request #310 from pegasystems/IH_analysis_improvements

Improved method signatures and examples
pegasystems · Jan 3, 2025 · 5f6a5c6 · 5f6a5c6
2 parents 2157a90 + 7f8bc3a
commit 5f6a5c6
Show file tree

Hide file tree

Showing 5 changed files with 292 additions and 125 deletions.
diff --git a/examples/ih/Conversion_Reporting.ipynb b/examples/ih/Conversion_Reporting.ipynb
@@ -33,17 +33,12 @@
     "from pathlib import Path\n",
     "import polars as pl\n",
     "\n",
-    "ih_export_file = Path(\n",
-    "    \"./Data-pxStrategyResult_InteractionFiles_20241213T091932_GMT.zip\"\n",
-    ")\n",
     "\n",
-    "if not ih_export_file.exists():\n",
-    "    ih = IH.from_mock_data()\n",
-    "else:\n",
-    "    ih = IH.from_ds_export(\n",
-    "        ih_export_file,\n",
-    "        query=pl.col.ExperimentGroup.is_not_null() & (pl.col.ExperimentGroup != \"\"),\n",
-    "    )\n",
+    "ih = IH.from_mock_data()\n",
+    "# ih = IH.from_ds_export(\n",
+    "#         \"./Data-pxStrategyResult_InteractionFiles_20241213T091932_GMT.zip\",\n",
+    "#     query=pl.col.ExperimentGroup.is_not_null() & (pl.col.ExperimentGroup != \"\"),\n",
+    "# )\n",
     "\n",
     "ih.aggregates.summary_success_rates(by=[\"ExperimentGroup\", \"Channel\"]).drop(\n",
     "    \"Outcomes\"\n",
@@ -79,7 +74,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "ih.plot.success_rates_tree_map(metric=\"Conversion\")\n"
+    "ih.plot.success_rate_tree_map(metric=\"Conversion\")\n"
    ]
   },
   {
@@ -97,20 +92,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "ih.plot.success_rates_trend_bar(\n",
-    "    metric=\"Conversion\",\n",
-    "    condition=\"ExperimentGroup\",\n",
-    "    every=\"1w\",\n",
-    ")\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "ih.plot.success_rates_trend(metric=\"Conversion\", every=\"1d\")"
+    "ih.plot.success_rate(metric=\"Conversion\", every=\"1d\")"
    ]
   },
   {
@@ -139,8 +121,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "ih.plot.success_rates_trend(\n",
-    "    by=\"Channel\"\n",
+    "ih.plot.success_rate(\n",
+    "    facet=\"Channel\"\n",
     ")"
    ]
   }
@@ -160,7 +142,8 @@
    "mimetype": "text/x-python",
    "name": "python",
    "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3"
+   "pygments_lexer": "ipython3",
+   "version": "3.12.3"
   }
  },
  "nbformat": 4,

diff --git a/examples/ih/Example_IH_Analysis.ipynb b/examples/ih/Example_IH_Analysis.ipynb
@@ -32,13 +32,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [],
    "source": [
-    "ih = IH.from_ds_export(\n",
-    "    \"../../data/Data-pxStrategyResult_pxInteractionHistory_20210101T010000_GMT.zip\"\n",
-    ")"
+    "# ih = IH.from_ds_export(\n",
+    "#     \"../../data/Data-pxStrategyResult_pxInteractionHistory_20210101T010000_GMT.zip\"\n",
+    "# )\n",
+    "ih = IH.from_mock_data()"
    ]
   },
   {
@@ -57,6 +58,11 @@
     "ih.data.head().collect()"
    ]
   },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": []
+  },
   {
    "cell_type": "markdown",
    "metadata": {},
@@ -115,7 +121,7 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "ih.plot.response_counts(every=\"1d\")"
+    "ih.plot.response_count(every=\"1d\")"
    ]
   },
   {
@@ -131,8 +137,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "ih.plot.response_counts(\n",
-    "    by=\"Channel\",\n",
+    "ih.plot.response_count(\n",
+    "    facet=\"Channel\",\n",
     "    query=pl.col.Channel != \"\",\n",
     ")"
    ]
@@ -159,8 +165,8 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "ih.plot.success_rates_trend(\n",
-    "    by=\"Channel\", query=pl.col.Channel.is_not_null() & (pl.col.Channel != \"\")\n",
+    "ih.plot.success_rate(\n",
+    "    facet=\"Channel\", query=pl.col.Channel.is_not_null() & (pl.col.Channel != \"\")\n",
     ")"
    ]
   },
@@ -218,10 +224,9 @@
     "    .to_list()\n",
     "    for c in channels\n",
     "]\n",
-    "\n",
     "fig = ff.create_distplot(plot_data, group_labels=channels, show_hist=False)\n",
-    "fig.update_layout(title=\"Propensity Distribution\")\n",
-    "fig"
+    "fig.update_layout(title=\"Propensity Distribution\", yaxis=dict(showticklabels=False))\n",
+    "fig\n"
    ]
   },
   {
@@ -250,23 +255,44 @@
     "    .to_list()\n",
     "    if c is not None and c != \"\"\n",
     "]\n",
-    "plot_data=ih.data.filter(pl.col.OutcomeTime.is_not_null()).group_by(\"InteractionID\").agg(\n",
-    "    [pl.col.OutcomeTime.min().alias(\"Decision_Time\")]+\n",
-    "    [pl.col.OutcomeTime.filter(pl.col.Outcome == o).max().alias(o) for o in outcomes],\n",
-    ").collect().unpivot(\n",
-    "    index=[\"InteractionID\", \"Decision_Time\"],\n",
-    "    variable_name=\"Outcome\",\n",
-    "    value_name=\"Time\",\n",
-    ").with_columns(\n",
-    "    Duration = (pl.col.Time - pl.col.Decision_Time).dt.total_seconds()\n",
-    ").filter(pl.col.Duration > 0)\n",
+    "plot_data = (\n",
+    "    ih.data.filter(pl.col.OutcomeTime.is_not_null())\n",
+    "    .group_by(\"InteractionID\")\n",
+    "    .agg(\n",
+    "        [pl.col.OutcomeTime.min().alias(\"Decision_Time\")]\n",
+    "        + [\n",
+    "            pl.col.OutcomeTime.filter(pl.col.Outcome == o).max().alias(o)\n",
+    "            for o in outcomes\n",
+    "        ],\n",
+    "    )\n",
+    "    .collect()\n",
+    "    .unpivot(\n",
+    "        index=[\"InteractionID\", \"Decision_Time\"],\n",
+    "        variable_name=\"Outcome\",\n",
+    "        value_name=\"Time\",\n",
+    "    )\n",
+    "    .with_columns(Duration=(pl.col.Time - pl.col.Decision_Time).dt.total_seconds())\n",
+    "    .filter(pl.col.Duration > 0)\n",
+    ")\n",
+    "\n",
+    "ordered_outcomes = (\n",
+    "    plot_data.group_by(\"Outcome\")\n",
+    "    .agg(Duration=pl.col(\"Duration\").median())\n",
+    "    .sort(\"Duration\")[\"Outcome\"]\n",
+    "    .to_list()\n",
+    ")\n",
+    "\n",
     "fig = px.box(\n",
     "    plot_data,\n",
     "    x=\"Duration\",\n",
     "    y=\"Outcome\",\n",
     "    color=\"Outcome\",\n",
-    "    template=\"pega\"\n",
+    "    template=\"pega\",\n",
+    "    category_orders={\"Outcome\": ordered_outcomes},\n",
+    "    points=False,\n",
+    "    title=\"Duration of Responses\"\n",
     ")\n",
+    "fig.update_layout(xaxis_title=\"Duration (seconds)\")\n",
     "fig"
    ]
   }
@@ -286,7 +312,8 @@
    "mimetype": "text/x-python",
    "name": "python",
    "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3"
+   "pygments_lexer": "ipython3",
+   "version": "3.12.3"
   }
  },
  "nbformat": 4,