Deploy from GitHub Actions

analysis.py

@@ -3001,3 +3001,249 @@ def plot_sector_line_charts(
     plot_data.insert(0, "parameter", parameter)
 
     return fig, param_data, plot_data
+
+
+@timer(include_params=True)
+def plot_sector_box_charts(
+    df: pd.DataFrame,
+    parameter: str,
+    color_scale: list[str] = COLOR_SCALE,
+) -> tuple[Figure, pd.DataFrame, pd.DataFrame]:
+    """
+    Create box plots showing the distribution of parameter values by sector and year.
+
+    Parameters:
+    -----------
+    df : pd.DataFrame
+        Input dataframe
+    parameter : str
+        Parameter to plot
+    color_scale : list[str]
+        List of colors to use for sectors
+
+    Returns:
+    --------
+    tuple[Figure, pd.DataFrame, pd.DataFrame]
+        - Figure: Matplotlib figure containing the box plots
+        - DataFrame: Filtered dataframe containing the raw data used in the plot
+        - DataFrame: Contains the plotted data points: mean, median, and quartiles
+    """
+    # Define consistent colors for styling
+    GREY10 = "#1a1a1a"
+    GREY30 = "#4d4d4d"
+    GREY40 = "#666666"
+    GREY75 = "#bfbfbf"
+    GREY91 = "#e8e8e8"
+
+    # Filter data for parameter
+    param_data = df[df["Org_Analyte_Name"] == parameter].copy()
+
+    # For Salinity, exclude Fresh Water Lakes
+    if parameter == "Salinity":
+        param_data = param_data[param_data["Sector"] != "Freshwater Lakes"]
+
+    # Calculate year and prepare data
+    param_data["Reporting_Year"] = param_data["Activity_Start_Date_Time"].dt.year
+    sectors = sorted(param_data["Sector"].unique())
+    years = sorted(param_data["Reporting_Year"].unique())
+
+    # Determine if log scale should be used
+    use_log_scale = parameter in [
+        "Turbidity",
+        "Fecal Coliform (MPN)",
+        "Total Nitrogen",
+        "Total Phosphorus",
+    ]
+
+    # Create figure with single column layout - increased width from 8 to 12
+    fig = plt.figure(figsize=(15, 2.5 * len(sectors)))
+
+    # Create box plots
+    for idx, sector in enumerate(sectors):
+        ax = plt.subplot(len(sectors), 1, idx + 1)
+        sector_data = param_data[param_data["Sector"] == sector]
+
+        bp = ax.boxplot(
+            [
+                sector_data[sector_data["Reporting_Year"] == year]["Org_Result_Value"]
+                for year in years
+            ],
+            labels=years,  # type: ignore
+            patch_artist=True,
+            medianprops=dict(color="black"),
+            flierprops=dict(
+                marker="o",
+                markerfacecolor=color_scale[idx],
+                alpha=0.5,
+                markersize=4,
+            ),
+            boxprops=dict(facecolor=color_scale[idx], alpha=0.6),
+            widths=0.6,
+            positions=range(len(years)),
+        )
+
+        # Set proper x-axis limits with padding
+        ax.set_xlim(-0.5, len(years) - 0.5)
+
+        ax.set_title(sector, pad=10, fontsize=10, color=GREY30)
+
+        if use_log_scale:
+            ax.set_yscale("log")
+
+        # Customize appearance
+        ax.grid(True, axis="y", alpha=0.15, linestyle="-", color="gray")
+        ax.spines["top"].set_visible(False)
+        ax.spines["right"].set_visible(False)
+        ax.spines["left"].set_visible(False)
+        ax.spines["bottom"].set_color(GREY40)
+        ax.spines["bottom"].set_linewidth(0.5)
+
+        # Customize tick parameters
+        ax.tick_params(axis="both", which="both", length=0, colors=GREY40)
+
+        # Update x-axis tick handling - removed rotation
+        if len(years) > 8:
+            # Show roughly every nth tick to get 8 or fewer ticks
+            n = len(years) // 8 + 1
+            visible_ticks = range(0, len(years), n)
+            visible_years = years[::n]
+            ax.set_xticks(visible_ticks)
+            ax.set_xticklabels(visible_years, ha="center", weight=500, color=GREY40)
+        else:
+            ax.set_xticks(range(len(years)))
+            ax.set_xticklabels(years, ha="center", weight=500, color=GREY40)
+
+    # Add overall title
+    fig.suptitle(
+        f"{parameter} Distribution by Sector", fontsize=14, y=1.02, color=GREY30
+    )
+
+    # Adjust layout - removed bottom adjustment since we no longer have rotated labels
+    plt.tight_layout()
+    plt.subplots_adjust(hspace=0.4)
+
+    # Create stats DataFrame to store box plot statistics
+    stats_data = []
+    for sector in sectors:
+        sector_data = param_data[param_data["Sector"] == sector]
+        for year in years:
+            year_data = sector_data[sector_data["Reporting_Year"] == year][
+                "Org_Result_Value"
+            ]
+            if not year_data.empty:
+                stats = {
+                    "Sector": sector,
+                    "Reporting_Year": year,
+                    "mean": year_data.mean(),
+                    "median": year_data.median(),
+                    "q1": year_data.quantile(0.25),
+                    "q3": year_data.quantile(0.75),
+                    "min": year_data.min(),
+                    "max": year_data.max(),
+                    "count": len(year_data),
+                }
+                stats_data.append(stats)
+
+    # Create stats DataFrame and add parameter column
+    stats_df = pd.DataFrame(stats_data)
+    stats_df.insert(0, "parameter", parameter)
+
+    return fig, param_data, stats_df
+
+
+@timer(include_params=True)
+def plot_sector_heatmap(
+    df: pd.DataFrame,
+    parameter: str,
+    show_values: bool = False,
+) -> tuple[Figure, pd.DataFrame, pd.DataFrame]:
+    """
+    Create a heatmap showing annual means by sector and year.
+
+    Parameters:
+    -----------
+    df : pd.DataFrame
+        Input dataframe
+    parameter : str
+        Name of the parameter to plot
+    show_values : bool, default=False
+        Whether to display mean values inside each cell
+
+    Returns:
+    --------
+    tuple[Figure, pd.DataFrame, pd.DataFrame]
+        - Figure: Matplotlib figure containing the heatmap
+        - DataFrame: Filtered dataframe containing the raw data used in the plot
+        - DataFrame: Contains the plotted data points: mean values for each sector and year
+    """
+    # Filter data for selected parameter
+    param_data = df[df["Org_Analyte_Name"] == parameter].copy()
+
+    # For Salinity, exclude Fresh Water Lakes
+    if parameter == "Salinity":
+        param_data = param_data[param_data["Sector"] != "Fresh Water Lakes"]
+
+    # Calculate annual means
+    plot_data = (
+        param_data.groupby(["Reporting_Year", "Sector"], observed=True)[
+            "Org_Result_Value"
+        ]
+        .mean()
+        .reset_index()
+        .pivot(index="Sector", columns="Reporting_Year", values="Org_Result_Value")
+    )
+
+    # Create figure with extra space at bottom for colorbar
+    fig, ax = plt.subplots(figsize=(12, len(plot_data) * 0.8))
+
+    # Create heatmap with small gaps between cells
+    im = ax.imshow(plot_data, aspect="auto", cmap="YlOrRd")
+
+    # Customize appearance
+    ax.set_xticks(np.arange(len(plot_data.columns)))
+    ax.set_yticks(np.arange(len(plot_data.index)))
+    ax.set_xticklabels(plot_data.columns)
+    ax.set_yticklabels(plot_data.index)
+
+    # Remove all spines
+    ax.spines["top"].set_visible(False)
+    ax.spines["right"].set_visible(False)
+    ax.spines["left"].set_visible(False)
+    ax.spines["bottom"].set_visible(False)
+
+    # Remove all tick marks but keep labels
+    ax.tick_params(axis="both", which="both", length=0)
+
+    # Add small gaps between cells
+    ax.set_xticks(np.arange(plot_data.shape[1] + 1) - 0.5, minor=True)
+    ax.set_yticks(np.arange(plot_data.shape[0] + 1) - 0.5, minor=True)
+    ax.grid(which="minor", color="w", linestyle="-", linewidth=2)
+
+    # Set x-axis labels horizontal
+    plt.setp(ax.get_xticklabels(), rotation=0)
+
+    # Add value annotations if requested
+    if show_values:
+        for i in range(len(plot_data.index)):
+            for j in range(len(plot_data.columns)):
+                value = plot_data.iloc[i, j]
+                if not pd.isna(value):
+                    text = f"{value:.1f}"
+                    ax.text(j, i, text, ha="center", va="center", color="black")
+
+    # Add colorbar at the bottom with reduced padding and no border
+    cbar = ax.figure.colorbar(im, ax=ax, orientation="horizontal", pad=0.1)  # type: ignore
+    unit = param_data["Org_Result_Unit"].iloc[0] if not param_data.empty else ""
+    cbar.ax.set_xlabel(f"Mean ({unit})")
+    cbar.outline.set_visible(False)  # type: ignore
+
+    # Set title
+    ax.set_title(parameter)
+
+    plt.tight_layout()
+
+    # Reset index to make Sector a column and add parameter column
+    plot_data = plot_data.reset_index()
+    plot_data.insert(0, "parameter", parameter)
+
+    return fig, param_data, plot_data

ui/pages/sector_compare.py

@@ -3,7 +3,11 @@ import io
 import pandas as pd
 import streamlit as st
 
-from analysis import plot_sector_line_charts
+from analysis import (
+    plot_sector_box_charts,
+    plot_sector_heatmap,
+    plot_sector_line_charts,
+)
 from components import render_filtered_data_preview
 from dashboard_analytics import log_visit
 
@@ -11,10 +15,35 @@ st.title("Sector Comparison Charts")
 log_visit("Sector Comparison Charts")
 
 # Get data from session state
-raw_df = st.session_state.data["raw_df"]
+raw_df = st.session_state.data["raw_df"].copy()
 raw_df["Date"] = pd.to_datetime(raw_df["Activity_Start_Date_Time"]).dt.date
 
 # Create sidebar controls
+chart_type = st.sidebar.radio(
+    "Chart Type:",
+    options=["Line Charts", "Box and Whisker", "Heatmap"],
+    help="Select the type of chart to display.",
+    key="sector_compare_chart_type",
+)
+
+# Only show relevant controls based on chart type
+if chart_type in ["Line Charts", "Heatmap"]:
+    st.sidebar.markdown("#### Chart Specific Options")
+if chart_type == "Line Charts":
+    show_sem = st.sidebar.checkbox(
+        "Show Standard Error",
+        value=True,
+        help="Display standard error bands around the mean lines",
+        key="sector_compare_show_sem",
+    )
+elif chart_type == "Heatmap":
+    show_values = st.sidebar.checkbox(
+        "Show Means",
+        value=False,
+        help="Display mean values inside each cell",
+        key="sector_compare_show_values",
+    )
+
 selected_parameter = st.sidebar.selectbox(
     "Parameter:",
     options=sorted(raw_df["Org_Analyte_Name"].unique()),
@@ -25,14 +54,6 @@ selected_parameter = st.sidebar.selectbox(
     key="sector_compare_parameter_select",
 )
 
-# Add show_sem checkbox
-show_sem = st.sidebar.checkbox(
-    "Show Standard Error",
-    value=True,
-    help="Display standard error bands around the mean lines",
-    key="sector_compare_show_sem",
-)
-
 # Add year filter slider
 years = sorted(raw_df["Activity_Start_Date_Time"].dt.year.unique())
 default_min_year = min(years)
@@ -54,15 +75,32 @@ filtered_raw_df = raw_df[
 ]
 
 try:
-    fig, param_data, plot_data = plot_sector_line_charts(
-        filtered_raw_df, selected_parameter, show_sem=show_sem
-    )
-    st.pyplot(fig)
+    if chart_type == "Line Charts":
+        # Line charts
+        fig, param_data, plot_data = plot_sector_line_charts(
+            filtered_raw_df, selected_parameter, show_sem=show_sem
+        )
+        st.pyplot(fig)
 
-    fig, _, _ = plot_sector_line_charts(
-        filtered_raw_df, selected_parameter, show_sem=show_sem, panel_chart=True
-    )
-    st.pyplot(fig)
+        fig, _, _ = plot_sector_line_charts(
+            filtered_raw_df, selected_parameter, show_sem=show_sem, panel_chart=True
+        )
+        st.pyplot(fig)
+    elif chart_type == "Box and Whisker":
+        # Add box plots
+        st.subheader("Distribution by Sector and Year")
+        fig, param_data, plot_data = plot_sector_box_charts(
+            filtered_raw_df, selected_parameter
+        )
+        st.pyplot(fig)
+    else:  # Heatmap
+        # Create heatmap
+        fig, param_data, plot_data = plot_sector_heatmap(
+            filtered_raw_df,
+            selected_parameter,
+            show_values=show_values if "show_values" in locals() else False,
+        )
+        st.pyplot(fig)
 
     # Add Chart Data expander
     with st.expander("Chart Data"):