Target app refactor (posit-dev#52)

* Remove conditional panel

* Add data tab with value boxes

Author: Gordon Shotwell

apps/target-app/app.py

@@ -1,5 +1,6 @@
+from numpy import char, place
 from shiny.express import ui, input, render
-from shiny import render_plot, req, reactive
+from shiny import reactive
 import pandas as pd
 from pathlib import Path
 from plots import (
@@ -11,45 +12,33 @@
 import faicons as fa
 import io
 from shinywidgets import render_plotly
+import time
 
 file_path = Path(__file__).parent / "simulated-data.csv"
 
-ui.page_opts(title="Monitoring")
 df = pd.read_csv(file_path, dtype={"sub_account": str})
 df["date"] = pd.to_datetime(df["date"], errors="coerce")
 
 
-@reactive.calc
-def monitor_sampled_data() -> pd.DataFrame:
-    start_date, end_date = input.dates()
-    start_date = pd.to_datetime(start_date)
-    end_date = pd.to_datetime(end_date)
-    df_value = df
-    out = df_value[
-        (df_value["date"] > start_date) & (df_value["date"] <= end_date)
-    ].sample(n=input.sample(), replace=True)
-    return out
-
-
 @reactive.calc()
-def monitor_filtered_data() -> pd.DataFrame:
-    sample_df = monitor_sampled_data()
-    sample_df = sample_df.loc[sample_df["account"] == input.account()]
-    return sample_df.reset_index(drop=True)
+def account_data():
+    return df[
+        (df["account"] == input.account()) & (df["sub_account"] == input.sub_account())
+    ]
 
 
 @reactive.calc()
-def training_data():
-    return df[
-        (df["account"] == input.account()) & (df["sub_account"] == input.sub_account())
+def character_filter():
+    return account_data()[
+        (account_data()["text"].str.len() >= input.chars()[0])
+        & (account_data()["text"].str.len() <= input.chars()[1])
     ]
 
 
 @reactive.effect
 @reactive.event(input.reset)
 def reset_vals():
-    ui.update_date_range("dates", start="2023-01-01", end="2023-04-01")
-    ui.update_numeric("sample", value=10000)
+    ui.update_slider("chars", value=[500, 6000])
 
 
 with ui.sidebar():
@@ -71,56 +60,70 @@ def sub_selector():
         choices = choice_data["sub_account"].unique().tolist()
         ui.input_select("sub_account", "Sub Account", choices=choices)
 
-    with ui.panel_conditional("input.tabs !== 'Training Dashboard'"):
-        ui.input_date_range(
-            "dates",
-            "Dates",
-            start="2023-01-01",
-            end="2023-04-01",
+    with ui.tooltip(id="btn_tooltip", placement="right"):
+        ui.input_slider(
+            "chars",
+            "Text length",
+            min=0,
+            max=df["text"].str.len().max(),
+            value=[500, 6000],
         )
-        ui.input_numeric("sample", "Sample Size", value=10000, step=5000)
-        ui.input_action_button("reset", "Reset Values", class_="btn-primary")
-
-
-with ui.navset_bar(id="tabs", title="Monitoring"):
-    with ui.nav_panel("Training Dashboard"):
-        with ui.layout_columns():
-            with ui.card():
-                ui.card_header("Model Metrics")
-
-                @render_plotly
-                def metric():
-                    if input.metric() == "ROC Curve":
-                        return plot_auc_curve(
-                            training_data(), "is_electronics", "training_score"
-                        )
-                    else:
-                        return plot_precision_recall_curve(
-                            training_data(), "is_electronics", "training_score"
-                        )
+        "The number of characters in the text"
+
+    ui.input_action_button("reset", "Reset Values", class_="btn-primary")
+
+with ui.nav_panel("Training Dashboard"):
+    with ui.layout_columns():
+        with ui.card():
+            ui.card_header("Model Metrics")
 
             @render_plotly
             def metric():
-                df_value = df()
-                df_filtered = df_value[df_value["account"] == input.account()]
                 if input.metric() == "ROC Curve":
                     return plot_auc_curve(
-                        df_filtered, "is_electronics", "training_score"
+                        character_filter(), "is_electronics", "training_score"
                     )
                 else:
                     return plot_precision_recall_curve(
-                        df_filtered, "is_electronics", "training_score"
+                        character_filter(), "is_electronics", "training_score"
                     )
 
-                @render_plotly
-                def score_dist():
-                    return plot_score_distribution(training_data())
+            ui.input_select(
+                "metric", "Metric", choices=["ROC Curve", "Precision Recall"]
+            )
+
+        with ui.card():
+            ui.card_header("Model Scores")
 
             @render_plotly
             def score_dist():
-                df_value = df()
-                df_filtered = df_value[df_value["account"] == input.account()]
-                return plot_score_distribution(df_filtered)
+                return plot_score_distribution(character_filter())
+
+
+vb_theme = "bg-blue"
+
+with ui.nav_panel("Data"):
+    with ui.layout_columns():
+        with ui.value_box(theme=vb_theme):
+            "Number of records"
+
+            @render.text
+            def data_count():
+                return str(character_filter().shape[0])
+
+        with ui.value_box(theme=vb_theme):
+            "Mean Score"
+
+            @render.text
+            def mean_score(theme=vb_theme):
+                return f"{character_filter()['training_score'].mean() * 100:.2f}%"
+
+        with ui.value_box(theme=vb_theme):
+            "Mean Text Length"
+
+            @render.text
+            def mean_text_length():
+                return f"{character_filter()['text'].str.len().mean():.2f} characters"
 
     with ui.card(full_screen=True):
         with ui.card_header():
@@ -134,27 +137,10 @@ def score_dist():
                 @render.download(filename=lambda: f"{input.account()}_scores.csv")
                 def download_data():
                     with io.BytesIO() as buf:
-                        filtered_data().to_csv(buf, index=False)
+                        account_data().to_csv(buf, index=False)
                         buf.seek(0)
                         yield buf.getvalue()
 
         @render.data_frame
         def data_output():
-            return filtered_data().drop(columns=["text"])
-
-
-with ui.nav_panel("Model Monitoring"):
-    with ui.layout_columns():
-        with ui.card():
-            ui.card_header("API Response Time")
-
-            @render_plotly
-            def api_response():
-                return plot_api_response(filtered_data())
-
-        with ui.card():
-            ui.card_header("Production Scores")
-
-            @render_plotly
-            def prod_score_dist():
-                return plot_score_distribution(filtered_data())
+            return account_data().drop(columns=["text"])