feat: plot route count timeseries

src/transport_performance/gtfs/multi_validation.py

@@ -12,6 +12,8 @@
 import pandas as pd
 import folium
 from folium.plugins import FastMarkerCluster
+import plotly.express as px
+import plotly.graph_objs as go
 
 from transport_performance.gtfs.validation import GtfsInstance
 from transport_performance.utils.defence import (
@@ -717,3 +719,108 @@ def get_dates(self, return_range: bool = True) -> list:
         if return_range:
             return [min(sorted_dates), max(sorted_dates)]
         return sorted_dates
+
+    def _reformat_col_names(self, col_name: str, cap_all: bool = True):
+        """Convert a column name to a more readable format."""
+        parts = col_name.split("_")
+        for i, part in enumerate(parts):
+            part = list(part)
+            part[0] = part[0].upper()
+            parts[i] = "".join(part)
+            if not cap_all:
+                break
+        return " ".join(parts)
+
+    def _plot_core(
+        self,
+        df: pd.DataFrame,
+        count_col: str = "trip_count",
+        width: int = 1000,
+        height: int = 550,
+        title: str = None,
+        kwargs: dict = {},
+    ):
+        """Plot a timeseries for trip/route count."""
+        # defences
+        _type_defence(df, "df", pd.DataFrame)
+        _type_defence(count_col, "count_col", str)
+        _type_defence(width, "width", int)
+        _type_defence(height, "height", int)
+        _type_defence(title, "title", (str, type(None)))
+        _type_defence(kwargs, "kwargs", dict)
+        # preparation
+        LABEL_FORMAT = {
+            count_col: self._reformat_col_names(count_col),
+            "date": "Date",
+        }
+        PLOT_TITLE = {
+            "text": f"{self._reformat_col_names(count_col, False)} over time",
+            "x": 0.5,
+            "xanchor": "center",
+        }
+        if "route_type" in df.columns:
+            kwargs["color"] = "route_type"
+            LABEL_FORMAT["route_type"] = self._reformat_col_names("route_type")
+            PLOT_TITLE["text"] = PLOT_TITLE["text"] + " by route type"
+        kwargs["width"] = width
+        kwargs["height"] = height
+        if title:
+            PLOT_TITLE["title"] = title
+        # plotting
+        fig = px.line(df, x="date", y=count_col, labels=LABEL_FORMAT, **kwargs)
+        fig.update_layout(title=PLOT_TITLE)
+
+        return fig
+
+    def plot_routes(
+        self,
+        route_type: bool = True,
+        width: int = 1000,
+        height: int = 550,
+        title: str = None,
+        plotly_kwargs: dict = None,
+    ) -> go.Figure:
+        """Create a line plot of route counts over time.
+
+        Parameters
+        ----------
+        route_type : bool, optional
+            Whether or not to draw a line for each modality, by default True
+        width : int, optional
+            Plot width, by default 1000
+        height : int, optional
+            Plot height, by default 550
+        title : str, optional
+            Plot title, by default None
+        plotly_kwargs : dict, optional
+            Kwargs to pass to plotly.express.line, by default None
+
+        Returns
+        -------
+        go.Figure
+            The timerseries plot
+
+        """
+        # defences
+        _type_defence(route_type, "route_type", bool)
+        _type_defence(plotly_kwargs, "plotly_kwargs", (dict, type(None)))
+        if not plotly_kwargs:
+            plotly_kwargs = {}
+        # prepare data
+        data = self.summarise_routes().copy()
+        if not route_type:
+            data = (
+                data.drop("route_type", axis=1)
+                .groupby("date")
+                .sum()
+                .reset_index()
+            )
+        figure = self._plot_core(
+            data,
+            "route_count",
+            width=width,
+            height=height,
+            title=title,
+            kwargs=plotly_kwargs,
+        )
+        return figure