Simple scatter plot with altair

requirements.txt

@@ -1,4 +1,4 @@
 pandas
 duckdb
 streamlit
-plotly
+altair

src/streamlit_app.py

@@ -1,10 +1,13 @@
 import os
+
+import altair as alt
 import duckdb
 import streamlit as st
-import plotly.express as px
 
-# Configuration
+# Page configuration
 st.set_page_config(layout="wide")
+
+# Data source
 url = (
     "stocks.parquet"
     if os.getenv("APP_ENV") == "development"
@@ -14,10 +17,10 @@ url = (
 
 @st.cache_data
 def get_data(filters=None):
-    base_query = "SELECT * FROM read_parquet('{}')".format(url)
+    query = f"SELECT * FROM read_parquet('{url}')"
     if filters:
-        base_query += f" WHERE {' AND '.join(filters)}"
-    return duckdb.query(base_query + " ORDER BY date DESC").to_df()
+        query += " WHERE " + " AND ".join(filters)
+    return duckdb.query(query + " ORDER BY date DESC").to_df()
 
 
 @st.cache_data
@@ -31,24 +34,22 @@ def load_symbols():
     )
 
 
-# Initialize filters and chart settings
-if "filters" not in st.session_state:
-    st.session_state.update(
-        {
-            "date": None,
-            "symbols": [],
-            "min_open": 1.0,
-            "min_gap": None,
-            "min_run": None,
-            "min_vol": 10000.0,
-            "min_relvol": 5.0,
-            "chart_type": "scatter",
-            "x_axis": "open",
-            "y_axis": "change_pct",
-        }
-    )
-
-# Filters panel
+# Initialize session state
+defaults = {
+    "date": None,
+    "symbols": [],
+    "min_open": 1.0,
+    "min_gap": None,
+    "min_run": None,
+    "min_vol": 10000.0,
+    "min_relvol": 5.0,
+    "x_axis": "open",
+    "y_axis": "change_pct",
+}
+for key, value in defaults.items():
+    st.session_state.setdefault(key, value)
+
+# Sidebar filters
 with st.sidebar:
     st.session_state.date = st.date_input("Date", st.session_state.date)
     st.session_state.symbols = st.multiselect(
@@ -74,26 +75,28 @@ with st.sidebar:
         "Min rel volume", value=st.session_state.min_relvol
     )
 
-# Build query filters
-query_filters = []
-if st.session_state.date:
-    query_filters.append(f"date = '{st.session_state.date.strftime('%Y-%m-%d')}'")
-if st.session_state.symbols:
-    query_filters.append(f"symbol IN {tuple(st.session_state.symbols)}")
-if st.session_state.min_open:
-    query_filters.append(f"open >= {st.session_state.min_open}")
-if st.session_state.min_vol:
-    query_filters.append(f"volume >= {st.session_state.min_vol}")
-if st.session_state.min_relvol:
-    query_filters.append(f"relative_volume >= {st.session_state.min_relvol}")
-if st.session_state.min_gap:
-    query_filters.append(f"gap_pct >= {st.session_state.min_gap}")
-if st.session_state.min_run:
-    query_filters.append(f"run_pct >= {st.session_state.min_run}")
-
-# Get and display data
-df = get_data(query_filters if query_filters else None)
-
+# Construct query filters
+f = st.session_state
+filters = []
+if f.date:
+    filters.append(f"date = '{f.date.strftime('%Y-%m-%d')}'")
+if f.symbols:
+    filters.append(f"symbol IN {tuple(f.symbols)}")
+if f.min_open:
+    filters.append(f"open >= {f.min_open}")
+if f.min_vol:
+    filters.append(f"volume >= {f.min_vol}")
+if f.min_relvol:
+    filters.append(f"relative_volume >= {f.min_relvol}")
+if f.min_gap:
+    filters.append(f"gap_pct >= {f.min_gap}")
+if f.min_run:
+    filters.append(f"run_pct >= {f.min_run}")
+
+# Load data
+df = get_data(filters if filters else None)
+
+# UI
 if df.empty:
     st.info("No data found with current filters")
 else:
@@ -103,73 +106,46 @@ else:
         st.dataframe(df, use_container_width=True)
 
     with tab2:
-        # Chart configuration
-        cols = st.columns(3)
         numeric_cols = df.select_dtypes("number").columns.tolist()
 
-        with cols[0]:
-            st.session_state.chart_type = st.selectbox(
-                "Chart Type",
-                ["scatter", "line", "bar", "histogram"],
-                index=["scatter", "line", "bar", "histogram"].index(
-                    st.session_state.chart_type
-                ),
-            )
-        with cols[1]:
-            st.session_state.x_axis = st.selectbox(
+        col1, col2 = st.columns(2)
+        with col1:
+            x_axis = st.selectbox(
                 "X-axis",
                 numeric_cols,
                 index=(
-                    numeric_cols.index(st.session_state.x_axis)
-                    if st.session_state.x_axis in numeric_cols
+                    numeric_cols.index(f.get("x_axis", "open"))
+                    if f.get("x_axis", "open") in numeric_cols
                     else 0
                 ),
+                key="x_axis",
             )
-        with cols[2]:
-            # For histogram, we don't need Y-axis selection
-            if st.session_state.chart_type != "histogram":
-                st.session_state.y_axis = st.selectbox(
-                    "Y-axis",
-                    numeric_cols,
-                    index=(
-                        numeric_cols.index(st.session_state.y_axis)
-                        if st.session_state.y_axis in numeric_cols
-                        else 1
-                    ),
-                )
-
-        # Generate the appropriate chart
-        if st.session_state.chart_type == "scatter":
-            fig = px.scatter(
-                df,
-                x=st.session_state.x_axis,
-                y=st.session_state.y_axis,
-                hover_data=["symbol"],
-                title=f"{st.session_state.y_axis} vs {st.session_state.x_axis}",
-            )
-        elif st.session_state.chart_type == "line":
-            fig = px.line(
-                df,
-                x=st.session_state.x_axis,
-                y=st.session_state.y_axis,
-                hover_data=["symbol"],
-                title=f"{st.session_state.y_axis} over {st.session_state.x_axis}",
-            )
-        elif st.session_state.chart_type == "bar":
-            fig = px.bar(
-                df,
-                x=st.session_state.x_axis,
-                y=st.session_state.y_axis,
-                hover_data=["symbol"],
-                title=f"{st.session_state.y_axis} by {st.session_state.x_axis}",
+        with col2:
+            y_axis = st.selectbox(
+                "Y-axis",
+                numeric_cols,
+                index=(
+                    numeric_cols.index(f.get("y_axis", "change_pct"))
+                    if f.get("y_axis", "change_pct") in numeric_cols
+                    else 1
+                ),
+                key="y_axis",
             )
-        elif st.session_state.chart_type == "histogram":
-            fig = px.histogram(
-                df,
-                x=st.session_state.x_axis,
-                title=f"Distribution of {st.session_state.x_axis}",
+
+        # Altair scatter plot
+        chart = (
+            alt.Chart(df)
+            .mark_circle(size=60)
+            .encode(
+                x=alt.X(x_axis, title=x_axis.capitalize()),
+                y=alt.Y(y_axis, title=y_axis.capitalize()),
+                color=alt.Color("symbol:N", legend=None),
+                tooltip=["symbol", x_axis, y_axis],
             )
+            .interactive()
+            .properties(title=f"{y_axis} vs {x_axis}")
+        )
 
-        st.plotly_chart(fig, use_container_width=True)
+        st.altair_chart(chart, use_container_width=True)
 
-st.write(f"Results: {len(df)}")
+    st.write(f"Results: {len(df)}")