BayDAG Contribution #10: NMTF Person Available Periods

activitysim/abm/models/non_mandatory_tour_frequency.py

@@ -11,7 +11,10 @@
 import pandas as pd
 
 from activitysim.abm.models.util import annotate
-from activitysim.abm.models.util.overlap import person_max_window
+from activitysim.abm.models.util.overlap import (
+    person_max_window,
+    person_available_periods,
+)
 from activitysim.abm.models.util.school_escort_tours_trips import (
     recompute_tour_count_statistics,
 )
@@ -230,7 +233,11 @@ def non_mandatory_tour_frequency(
     # - preprocessor
     preprocessor_settings = model_settings.preprocessor
     if preprocessor_settings:
-        locals_dict = {"person_max_window": lambda x: person_max_window(state, x)}
+
+        locals_dict = {
+            "person_max_window": lambda x: person_max_window(state, x),
+            "person_available_periods": lambda x: person_available_periods(state, x),
+        }
 
         expressions.assign_columns(
             state,
@@ -324,6 +331,9 @@ def non_mandatory_tour_frequency(
 
         choices_list.append(choices)
 
+    # FIXME only want to keep actual purposes, adding cols in alts will mess this up
+    # this is complicated by canonical_ids calculated based on alts if not specified explicitly
+    # thus, adding column to input alts will change IDs and break estimation mode....
     del alternatives["tot_tours"]  # del tot_tours column we added above
 
     # The choice value 'non_mandatory_tour_frequency' assigned by interaction_simulate
@@ -418,13 +428,14 @@ def non_mandatory_tour_frequency(
     if estimator:
         # make sure they created the right tours
         survey_tours = estimation.manager.get_survey_table("tours").sort_index()
-        non_mandatory_survey_tours = survey_tours[
-            survey_tours.tour_category == "non_mandatory"
-        ]
-        assert len(non_mandatory_survey_tours) == len(non_mandatory_tours)
-        assert non_mandatory_survey_tours.index.equals(
-            non_mandatory_tours.sort_index().index
-        )
+        # FIXME below check needs to remove the pure-escort tours from the survey tours table
+        # non_mandatory_survey_tours = survey_tours[
+        #     survey_tours.tour_category == "non_mandatory"
+        # ]
+        # assert len(non_mandatory_survey_tours) == len(non_mandatory_tours)
+        # assert non_mandatory_survey_tours.index.equals(
+        #     non_mandatory_tours.sort_index().index
+        # )
 
         # make sure they created tours with the expected tour_ids
         columns = ["person_id", "household_id", "tour_type", "tour_category"]

activitysim/abm/models/util/overlap.py

@@ -252,3 +252,98 @@ def person_max_window(state: workflow.State, persons):
     max_window.index = persons.index
 
     return max_window
+
+
+def calculate_consecutive(array):
+    # Append zeros columns at either sides of counts
+    append1 = np.zeros((array.shape[0], 1), dtype=int)
+    array_ext = np.column_stack((append1, array, append1))
+
+    # Get start and stop indices with 1s as triggers
+    diffs = np.diff((array_ext == 1).astype(int), axis=1)
+    starts = np.argwhere(diffs == 1)
+    stops = np.argwhere(diffs == -1)
+
+    # Get intervals using differences between start and stop indices
+    intvs = stops[:, 1] - starts[:, 1]
+
+    # Store intervals as a 2D array for further vectorized ops to make.
+    c = np.bincount(starts[:, 0])
+    mask = np.arange(c.max()) < c[:, None]
+    intvs2D = mask.astype(float)
+    intvs2D[mask] = intvs
+
+    # Get max along each row as final output
+    out = intvs2D.max(1).astype(int)
+    return out
+
+
+def person_available_periods(
+    state: workflow.State, persons, start_bin=None, end_bin=None, continuous=False
+):
+    """
+    Returns the number of available time period bins foreach person in persons.
+    Can limit the calculation to include starting and/or ending bins.
+    Can return either the total number of available time bins with continuous = True,
+    or only the maximum
+
+    This is equivalent to person_max_window if no start/end bins provided and continous=True
+
+    time bins are inclusive, i.e. [start_bin, end_bin]
+
+    e.g.
+    available out of timetable has dummy first and last bins
+    available = [
+        [1,1,1,1,1,1,1,1,1,1,1,1],
+        [1,1,0,1,1,0,0,1,0,1,0,1],
+        #-,0,1,2,3,4,5,6,7,8,9,-  time bins
+    ]
+    returns:
+    for start_bin=None, end_bin=None, continuous=False: (10, 5)
+    for start_bin=None, end_bin=None, continuous=True: (10, 2)
+    for start_bin=5, end_bin=9, continuous=False: (5, 2)
+    for start_bin=5, end_bin=9, continuous=True: (5, 1)
+
+
+    Parameters
+    ----------
+    start_bin : (int) starting time bin to include starting from 0
+    end_bin : (int) ending time bin to include
+    continuous : (bool) count all available bins if false or just largest continuous run if True
+
+    Returns
+    -------
+    pd.Series of the number of available time bins indexed by person ID
+    """
+    timetable = state.get_injectable("timetable")
+
+    # ndarray with one row per person and one column per time period
+    # array value of 1 where free periods and 0 elsewhere
+    s = pd.Series(persons.index.values, index=persons.index)
+
+    # first and last bins are dummys in the time table
+    # so if you have 48 half hour time periods, shape is (len(persons), 50)
+    available = timetable.individually_available(s)
+
+    # Create a mask to exclude bins before the starting bin and after the ending bin
+    mask = np.ones(available.shape[1], dtype=bool)
+    mask[0] = False
+    mask[len(mask) - 1] = False
+    if start_bin is not None:
+        # +1 needed due to dummy first bin
+        mask[: start_bin + 1] = False
+    if end_bin is not None:
+        # +2 for dummy first bin and inclusive end_bin
+        mask[end_bin + 2 :] = False
+
+    # Apply the mask to the array
+    masked_array = available[:, mask]
+
+    # Calculate the number of available time periods for each person
+    availability = np.sum(masked_array, axis=1)
+
+    if continuous:
+        availability = calculate_consecutive(masked_array)
+
+    availability = pd.Series(availability, index=persons.index)
+    return availability