chore: Commit prior to context switch

notebooks/osm/01-visualise-osm.py

@@ -1,24 +1,25 @@
 """Visualise OMS spatial features."""
+# %%
 from pyprojroot import here
 import pandas as pd
 import geopandas as gpd
 from shapely import Point
+from typing import Union
 from transport_performance.osm import validate_osm as osmval
 
 # %%
 FIX_PTH = here("tests/data/newport-2023-06-13.osm.pbf")
 
 # %%
 # get some Node IDs to work with
-
-nodes = osmval.FindIds(FIX_PTH)
-nodes.get_feature_ids()
-some_IDs = nodes.id_dict["node_ids"][0:100]
+ids = osmval.FindIds(FIX_PTH)
+ids.get_feature_ids()
+node_IDs = ids.id_dict["node_ids"][0:1000]
 
 # %%
 # get the locations
 locs = osmval.FindLocations(FIX_PTH)
-some_coords = locs.check_locs_for_ids(some_IDs, "node")
+node_coords = locs.check_locs_for_ids(node_IDs, "node")["node"]
 # %%
 # get the contextual information in the tags, though note these
 # be blank for nodes
@@ -27,19 +28,74 @@
 # all of the found tags are blank. TODO: find a populated example.
 
 # %%
-all_coords = pd.DataFrame()
-for key, values in some_coords["node"].items():
-    coord_row = pd.DataFrame(values, index=[key])
-    all_coords = pd.concat([all_coords, coord_row])
 
-# %%
-all_coords["geometry"] = [
-    Point(xy) for xy in zip(all_coords["lon"], all_coords["lat"])
-]
+
+def _convert_osmdict_to_gdf(
+    _dict: dict,
+    feature_type: str = "node",
+    lon_colnm: str = "lon",
+    lat_colnm: str = "lat",
+    _crs: Union[str, int] = "epsg:4326",
+):
+    """Convert an OSM dictionary to a GDF.
+
+    Parameters
+    ----------
+    _dict: (dict)
+        Dictionary of ID: location / tags.
+    feature_type: str
+        The type of feature data contained in the dictionary. Defaults to
+        "node".
+    lon_colnm: (str)
+        Name of the longitude column.
+    lat_colnm: (str)
+        Name of the latitude column.
+    _crs: (Union[str, int], optional)
+        The CRS of the spatial features. Defaults to "epsg:4326".
+
+    Returns
+    -------
+    out_gdf: (gpd.GeoDataFrame)
+        A GeoDataFrame of the spatial features with ID mapped to index and
+        values mapped to columns.
+
+    """
+    out_gdf = pd.DataFrame()
+    out_row = pd.DataFrame()
+    for key, values in _dict.items():
+        if feature_type == "node":
+            out_row = pd.DataFrame(values, index=[key])
+        elif feature_type == "way":
+            # now we have a nested list of node dictionaries. These are node
+            # members of the way.
+            for i in values:
+                for k, j in i.items():
+                    mem_row = pd.DataFrame(j, index=[key, k])
+                    # if there are members of the parent ID, return multiindex
+                    ind = pd.MultiIndex.from_tuples(
+                        [(key, k)], names=["parent_id", "member_id"]
+                    )
+                    mem_row = pd.DataFrame(j, index=ind)
+                    out_row = pd.concat([out_row, mem_row])
+        # TODO - this line is behaving differently for feature type node versus
+        # way. needs figuring out.
+        out_gdf = pd.concat([out_gdf, out_row])
+
+    out_gdf["geometry"] = [
+        Point(xy) for xy in zip(out_gdf[lon_colnm], out_gdf[lat_colnm])
+    ]
+    out_gdf = gpd.GeoDataFrame(out_gdf, crs=_crs)
+    return out_gdf
+
 
 # %%
-coords_gdf = gpd.GeoDataFrame(all_coords, crs=4326)
-coords_gdf.explore()
+nodes_gdf = _convert_osmdict_to_gdf(node_coords)
+nodes_gdf.explore()
+
 
 # %%
+way_ids = ids.id_dict["way_ids"][0:400]
+way_coords = locs.check_locs_for_ids(way_ids, "way")["way"]
 # %%
+ways_gdf = _convert_osmdict_to_gdf(way_coords, "way", "x", "y")
+ways_gdf.explore()