Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Fix interpolation of trailing NAs for each trip #7

Merged
merged 2 commits into from
Mar 13, 2017
Merged

Fix interpolation of trailing NAs for each trip #7

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
4d1b2cb
Revise interpolator to ignore trailing NaNs for each trip and add tes…
pksohn Mar 13, 2017
a0e16b9
Merge branch 'edge_formatter' into interpolator
pksohn Mar 13, 2017
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 19 additions & 3 deletions urbanaccess/gtfs/network.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -257,15 +257,31 @@ def interpolatestoptimes(stop_times_df, calendar_selected_trips_df, day):
limit_direction='forward')

# Melt back into stacked format
interpolator['stop_sequence'] = interpolator.index
melted = pd.melt(interpolator, id_vars='stop_sequence')
interpolator['stop_sequence_merge'] = interpolator.index
melted = pd.melt(interpolator, id_vars='stop_sequence_merge')
melted.rename(columns={'value': 'departure_time_sec_interpolate'},
inplace=True)

# Get the last valid stop for each unique trip, to filter out trailing NaNs
last_valid_stop_series = pivot.apply(
lambda col: col.last_valid_index(), axis=0)
last_valid_stop_df = last_valid_stop_series.to_frame('last_valid_stop')

df_for_interpolation = df_for_interpolation.merge(last_valid_stop_df,
left_on='unique_trip_id',
right_index=True)
trailing = (df_for_interpolation.stop_sequence >
df_for_interpolation.last_valid_stop)

# Calculate a stop_sequence without trailing NaNs, to merge the correct
# interpolated times back in
df_for_interpolation['stop_sequence_merge'] = (
df_for_interpolation[~trailing]['stop_sequence'])

# Merge back into original index
df_for_interpolation.reset_index(inplace=True)
interpolated_df = pd.merge(df_for_interpolation, melted, 'left',
on=['stop_sequence', 'unique_trip_id'])
on=['stop_sequence_merge', 'unique_trip_id'])
interpolated_df.set_index('index', inplace=True)
interpolated_times = interpolated_df[['departure_time_sec_interpolate']]

Expand Down
30 changes: 18 additions & 12 deletions urbanaccess/tests/test_gtfs_network.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7,19 +7,21 @@
@pytest.fixture
def stop_times():
data = {
'unique_agency_id': ['citytrains'] * 20,
'unique_agency_id': ['citytrains'] * 25,
'trip_id': ['a', 'a', 'a', 'a', 'a',
'b', 'b', 'b', 'b', 'b',
'c', 'c', 'c', 'c', 'c',
'd', 'd', 'd', 'd', 'd'],
'stop_id': range(20),
'd', 'd', 'd', 'd', 'd',
'e', 'e', 'e', 'e', 'e'],
'stop_id': range(25),
'departure_time_sec': [1, 2, np.nan, np.nan, 5,
1, 2, 3, 4, np.nan,
np.nan, np.nan, 3, 4, np.nan,
1, 2, 3, 4, 5],
'stop_sequence': [1, 2, 3, 4, 5] * 4
1, 2, 3, 4, 5,
1, np.nan, 3, 4, np.nan],
'stop_sequence': [1, 2, 3, 4, 5] * 5
}
index = range(20)
index = range(25)

df = pd.DataFrame(data, index)
return df
Expand All @@ -28,10 +30,10 @@ def stop_times():
@pytest.fixture
def calendar():
data = {
'unique_agency_id': ['citytrains'] * 3,
'trip_id': ['a', 'b', 'c']
'unique_agency_id': ['citytrains'] * 4,
'trip_id': ['a', 'b', 'c', 'e']
}
index = range(3)
index = range(4)

df = pd.DataFrame(data, index)
return df
Expand Down Expand Up @@ -96,13 +98,17 @@ def test_interpolator(stop_times, calendar):

# trip 'c' should be interpolated
# no starting value, so first two times removed
# missing value at end should be equal to last existing value
# NaN values should be removed from start and end
assert df.loc[df.trip_id == 'c',
'departure_time_sec_interpolate'].tolist() == [3, 4, 4]
'departure_time_sec_interpolate'].tolist() == [3, 4]

# trip 'd' should be removed because it's not in the calendar df
assert len(df.loc[df.trip_id == 'd']) == 0

# trip 'e' should interpolate the second row but leave off the trailing NA
assert df.loc[df.trip_id == 'e',
'departure_time_sec_interpolate'].tolist() == [1, 2, 3, 4]


def test_edge_reformatter(stop_times_interpolated):
df = network.format_transit_net_edge(stop_times_interpolated)
Expand Down Expand Up @@ -132,4 +138,4 @@ def test_edge_reformatter(stop_times_interpolated):
assert df['unique_trip_id'][8] == stop_times_interpolated[
'unique_trip_id'][11] and \
df['unique_agency_id'][8] == stop_times_interpolated[
'unique_agency_id'][11]
'unique_agency_id'][11]