Merge branch 'main' into 53-paths-with-79

config/base.toml

@@ -38,10 +38,16 @@ optional_columns = [
 n_matches = 10 # What is the maximum number of NTS matches we want for each SPC household?
 
 [feasible_assignment]
-# detour factor when converting euclidian distance to actual travel distance
+# `actual_distance = distance * (1 + ((detour_factor - 1) * np.exp(-decay_rate * distance)))`
+#
+# `detour factor` when converting Euclidean distance to actual travel distance
 detour_factor = 1.56
-# decay rate when converting euclidian to travel distance (0.0001 is a good value)
-# actual_distance = distance * (1 + ((detour_factor - 1) * np.exp(-decay_rate * distance)))
+
+# `decay rate` is the inverse of the distance (in units of the data, e.g. metres) at which the
+# scaling from using the detour factor to Euclidean distance reduces by `exp(−1)`.
+#
+# 0.0001 is a good value for Leeds when units are metres, choice of decay_rate can be explored in an
+# [interactive plot](https://www.wolframalpha.com/input?i=plot+exp%28-0.0001x%29+from+x%3D0+to+x%3D50000)
 decay_rate = 0.0001
 
 [work_assignment]

scripts/0.1_run_osmox.py

@@ -39,7 +39,9 @@ def main(config_file):
             "-f",
             "geoparquet",
             "-crs",
-            # TODO: check if this can be specified as the output CRS
+            # For the distances to be accurate, needs to be same CRS as OSM data for the region.
+            # However, distances from osmox are currently not used in the pipeline so any CRS will work.
+            # In general, the CRS is transformed in the pipeline when this data is used.
             # See: https://github.com/arup-group/osmox/blob/82602d411374ebc9fd33443f8f7c9816b63715ec/docs/osmox_run.md#L35-L38
             "epsg:27700",
             "-l",

src/acbm/logger_config.py

@@ -1,11 +1,4 @@
 import logging
-from datetime import datetime
-
-
-def prepend_datetime(s: str, delimiter: str = "_") -> str:
-    current_date = datetime.now().strftime("%Y-%m-%d")
-    return f"{current_date}{delimiter}{s}"
-
 
 # # Configure the root logger
 # logging.basicConfig(

tests/test_assigning.py

@@ -1,7 +1,8 @@
+import numpy as np
 import pandas as pd
 import pytest
 
-from acbm.assigning.utils import _map_day_to_wkday_binary
+from acbm.assigning.utils import _adjust_distance, _map_day_to_wkday_binary
 
 # applying to a single value
 
@@ -32,3 +33,17 @@ def test_map_day_to_wkday_binary_df_invalid():
     df = pd.DataFrame({"day": [1, 2, 3, 4, 5, 6, 7, 8]})
     with pytest.raises(ValueError, match="Day should be numeric and in the range 1-7"):
         df["wkday"] = df["day"].apply(_map_day_to_wkday_binary)
+
+
+def test_adjust_distance():
+    assert np.isclose(_adjust_distance(0, 1.5, 0.1), 0)
+    assert np.isclose(_adjust_distance(10000, 1.56, 0.0001), 12060.125)
+    assert np.isclose(
+        _adjust_distance(10000, 1.56, 0.0001),
+        10000 * (1 + ((1.56 - 1) * np.exp(-0.0001 * 10000))),
+    )
+    assert np.isclose(_adjust_distance(50000, 1.8, 0.0002), 50001.81599)
+    assert np.isclose(
+        _adjust_distance(50000, 1.8, 0.0002),
+        50000 * (1 + ((1.8 - 1) * np.exp(-0.0002 * 50000))),
+    )