def test_triplegs_read_csv_crs_parameter(self):
file = os.path.join('tests', 'data', 'triplegs.csv')
crs = "EPSG:2056"
tpls = ti.read_triplegs_csv(file, sep=';', tz='utc', index_col="id")
assert tpls.crs is None
tpls = ti.read_triplegs_csv(file, sep=';', tz='utc', index_col="id", crs=crs)
assert tpls.crs == crs
def test_from_to_csv(self):
"""Test basic reading and writing functions."""
orig_file = os.path.join("tests", "data", "triplegs.csv")
mod_file = os.path.join("tests", "data", "triplegs_mod_columns.csv")
tmp_file = os.path.join("tests", "data", "triplegs_test_1.csv")
tpls = ti.read_triplegs_csv(orig_file,
sep=";",
tz="utc",
index_col="id")
column_mapping = {
"start_time": "started_at",
"end_time": "finished_at",
"tripleg": "geom"
}
mod_tpls = ti.read_triplegs_csv(mod_file,
sep=";",
columns=column_mapping,
index_col="id")
assert mod_tpls.equals(tpls)
tpls["started_at"] = tpls["started_at"].apply(
lambda d: d.isoformat().replace("+00:00", "Z"))
tpls["finished_at"] = tpls["finished_at"].apply(
lambda d: d.isoformat().replace("+00:00", "Z"))
columns = ["user_id", "started_at", "finished_at", "geom"]
tpls.as_triplegs.to_csv(tmp_file, sep=";", columns=columns)
assert filecmp.cmp(orig_file, tmp_file, shallow=False)
os.remove(tmp_file)
def test_triplegs_csv_index_col(self):
"""Test if `index_col` can be set."""
file = os.path.join('tests', 'data', 'triplegs.csv')
ind_name = 'id'
pfs = ti.read_triplegs_csv(file, sep=";", index_col=ind_name)
assert pfs.index.name == ind_name
pfs = ti.read_triplegs_csv(file, sep=";", index_col=None)
assert pfs.index.name is None
def test_set_index(self):
"""Test if `index_col` can be set."""
file = os.path.join("tests", "data", "triplegs.csv")
ind_name = "id"
pfs = ti.read_triplegs_csv(file, sep=";", index_col=ind_name)
assert pfs.index.name == ind_name
pfs = ti.read_triplegs_csv(file, sep=";", index_col=None)
assert pfs.index.name is None
def test_set_crs(self):
"""Test setting the crs when reading."""
file = os.path.join("tests", "data", "triplegs.csv")
crs = "EPSG:2056"
tpls = ti.read_triplegs_csv(file, sep=";", tz="utc", index_col="id")
assert tpls.crs is None
tpls = ti.read_triplegs_csv(file,
sep=";",
tz="utc",
index_col="id",
crs=crs)
assert tpls.crs == crs
def test_test_Douglas_Peucker_Algorithm_has_no_side_effects(self):
orig_file = 'tests/data/triplegs_with_too_many_points_test.csv'
tpls = ti.read_triplegs_csv(orig_file, sep=';')
tpls_copy = copy.deepcopy(tpls)
tpls_smoothed = smoothen_triplegs(tpls, epsilon=0.0001)
assert np.all(tpls == tpls_copy)
def test_filter_triplegs(self):
# read triplegs and area file
tpls_file = os.path.join('tests', 'data', 'geolife',
'geolife_triplegs.csv')
tpls = ti.read_triplegs_csv(tpls_file, tz='utc', index_col='id')
extent = gpd.read_file(
os.path.join('tests', 'data', 'area', 'tsinghua.geojson'))
# the projection needs to be defined: WGS84
tpls.crs = 'epsg:4326'
within_tl = tpls.as_triplegs.spatial_filter(areas=extent,
method="within",
re_project=True)
intersects_tl = tpls.as_triplegs.spatial_filter(areas=extent,
method="intersects",
re_project=True)
crosses_tl = tpls.as_triplegs.spatial_filter(areas=extent,
method="crosses",
re_project=True)
# the result obtained from ArcGIS
gis_within_num = 9
gis_intersects_num = 20
assert len(within_tl) == gis_within_num, "The within tripleg number should be the same as" + \
"the one from the result with ArcGIS"
assert len(intersects_tl) == gis_intersects_num, "The intersects tripleg number should be " + \
"the same as the one from the result with ArcGIS"
assert len(crosses_tl) == len(intersects_tl) - len(within_tl), "The crosses tripleg number" + \
"should equal the number of intersect triplegs minus the number of within triplegs"
def test_filter_triplegs(self):
"""Test if spatial_filter works for triplegs."""
# read triplegs and area file
tpls_file = os.path.join("tests", "data", "geolife",
"geolife_triplegs.csv")
tpls = ti.read_triplegs_csv(tpls_file, tz="utc", index_col="id")
extent = gpd.read_file(
os.path.join("tests", "data", "area", "tsinghua.geojson"))
# the projection needs to be defined: WGS84
tpls.crs = "epsg:4326"
within_tl = tpls.as_triplegs.spatial_filter(areas=extent,
method="within",
re_project=True)
intersects_tl = tpls.as_triplegs.spatial_filter(areas=extent,
method="intersects",
re_project=True)
crosses_tl = tpls.as_triplegs.spatial_filter(areas=extent,
method="crosses",
re_project=True)
# the result obtained from ArcGIS
gis_within_num = 9
gis_intersects_num = 20
assert len(within_tl) == gis_within_num, (
"The within tripleg number should be the same as" +
"the one from the result with ArcGIS")
assert len(intersects_tl) == gis_intersects_num, (
"The intersects tripleg number should be " +
"the same as the one from the result with ArcGIS")
assert len(crosses_tl) == len(intersects_tl) - len(within_tl), (
"The crosses tripleg number" +
"should equal the number of intersect triplegs minus the number of within triplegs"
)
def test_Douglas_Peucker_Algorithm_reduces_triplet_length(self):
def plot_line(ax, ob):
x, y = ob.xy
ax.plot(x,
y,
alpha=0.7,
linewidth=3,
solid_capstyle='round',
zorder=2)
pd.set_option('display.max_columns', 10)
pd.set_option('display.max_rows', 10)
orig_file = 'tests/data/triplegs_with_too_many_points_test.csv'
tpls = ti.read_triplegs_csv(orig_file, sep=';')
tpls_smoothed = smoothen_triplegs(tpls, epsilon=0.0001)
line1 = tpls.iloc[0].geom
line1_smoothed = tpls_smoothed.iloc[0].geom
line2 = tpls.iloc[1].geom
line2_smoothed = tpls_smoothed.iloc[1].geom
print(line1)
print(line1_smoothed)
assert line1.length == line1_smoothed.length
assert line2.length == line2_smoothed.length
assert len(line1.coords) == 10
assert len(line2.coords) == 7
assert len(line1_smoothed.coords) == 4
assert len(line2_smoothed.coords) == 3
def test_triplegs_from_to_csv(self):
orig_file = os.path.join('tests', 'data', 'triplegs.csv')
mod_file = os.path.join('tests', 'data', 'triplegs_mod_columns.csv')
tmp_file = os.path.join('tests', 'data', 'triplegs_test.csv')
tpls = ti.read_triplegs_csv(orig_file, sep=';', tz='utc', index_col="id")
column_mapping = {'start_time': 'started_at', 'end_time': 'finished_at', 'tripleg': 'geom'}
mod_tpls = ti.read_triplegs_csv(mod_file, sep=';', columns=column_mapping, index_col="id")
assert mod_tpls.equals(tpls)
tpls['started_at'] = tpls['started_at'].apply(lambda d: d.isoformat().replace('+00:00', 'Z'))
tpls['finished_at'] = tpls['finished_at'].apply(lambda d: d.isoformat().replace('+00:00', 'Z'))
columns = ['user_id', 'started_at', 'finished_at', 'geom']
tpls.as_triplegs.to_csv(tmp_file, sep=';', columns=columns)
assert filecmp.cmp(orig_file, tmp_file, shallow=False)
os.remove(tmp_file)
def test_simple_coarse_identification_geographic(self):
"""Asserts that a warning is thrown if data in non-WGS geographic coordinate systems."""
tpls_file = os.path.join('tests', 'data',
'triplegs_transport_mode_identification.csv')
tpls = ti.read_triplegs_csv(tpls_file, sep=';', index_col="id")
tpls_2 = tpls.set_crs(epsg=4326)
tpls_4 = tpls_2.to_crs(epsg=4269)
with pytest.warns(UserWarning):
tpls_4.as_triplegs.predict_transport_mode(method='simple-coarse')
def test_check_empty_dataframe(self):
"""Assert that the method does not work for empty DataFrames
(but that the rest works fine, e.g., method signature).
"""
tpls_file = os.path.join('tests', 'data', 'triplegs_transport_mode_identification.csv')
tpls = ti.read_triplegs_csv(tpls_file, sep=';', index_col="id")
empty_frame = tpls[0:0]
with pytest.raises(AssertionError):
empty_frame.as_triplegs.predict_transport_mode(method='simple-coarse')
def test_check_empty_dataframe(self):
"""Assert that the method does not work for empty DataFrames."""
tpls_file = os.path.join("tests", "data",
"triplegs_transport_mode_identification.csv")
tpls = ti.read_triplegs_csv(tpls_file, sep=";", index_col="id")
empty_frame = tpls[0:0]
with pytest.raises(AssertionError):
empty_frame.as_triplegs.predict_transport_mode(
method="simple-coarse")
def test_simple_coarse_identification_geographic(self):
"""Asserts the correct behaviour with data in geographic coordinate systems."""
tpls_file = os.path.join('tests', 'data', 'triplegs_transport_mode_identification.csv')
tpls = ti.read_triplegs_csv(tpls_file, sep=';', index_col="id")
tpls_2 = tpls.set_crs(epsg=4326)
tpls_4 = tpls_2.to_crs(epsg=4269)
with pytest.raises(UserWarning,
match='Your data is in a geographic coordinate system, length calculation fails'):
tpls_4.as_triplegs.predict_transport_mode(method='simple-coarse')
def test_triplegs_from_gpd(self):
gdf = gpd.read_file(os.path.join('tests', 'data', 'triplegs.geojson'))
gdf.set_index('id', inplace=True)
tpls_from_gpd = ti.io.from_geopandas.triplegs_from_gpd(gdf, user_id='User', geom='geometry', tz='utc')
tpls_file = os.path.join('tests', 'data', 'triplegs.csv')
tpls_from_csv = ti.read_triplegs_csv(tpls_file, sep=';', tz='utc', index_col='id')
pd.testing.assert_frame_equal(tpls_from_gpd, tpls_from_csv, check_exact=False)
def test_triplegs_plot(self):
tmp_file = 'tests/data/triplegs_plot.png'
pfs = ti.read_positionfixes_csv('tests/data/positionfixes.csv',
sep=';')
tpls = ti.read_triplegs_csv('tests/data/triplegs.csv', sep=';')
tpls.as_triplegs.plot(out_filename=tmp_file,
positionfixes=pfs,
plot_osm=False)
assert os.path.exists(tmp_file)
os.remove(tmp_file)
def test_trajectory_distance_via_accessor_x(self):
tpls = ti.read_triplegs_csv(
os.path.join('tests', 'data', 'geolife', 'geolife_triplegs.csv'))
D_single = tpls.iloc[0:4].as_triplegs.similarity(dist_metric='dtw',
n_jobs=1)
D_multi = tpls.iloc[0:4].as_triplegs.similarity(dist_metric='dtw',
n_jobs=4)
assert np.isclose(np.sum(np.abs(D_single - D_multi)), 0)
def test_simple_coarse_identification_projected(self):
"""Asserts the correct behaviour with data in projected coordinate systems."""
tpls_file = os.path.join('tests', 'data', 'triplegs_transport_mode_identification.csv')
tpls = ti.read_triplegs_csv(tpls_file, sep=';', index_col="id")
tpls_2 = tpls.set_crs(epsg=4326)
tpls_3 = tpls_2.to_crs(epsg=2056)
tpls_transport_mode_3 = tpls_3.as_triplegs.predict_transport_mode(method='simple-coarse')
assert tpls_transport_mode_3.iloc[0]['mode'] == 'slow_mobility'
assert tpls_transport_mode_3.iloc[1]['mode'] == 'motorized_mobility'
assert tpls_transport_mode_3.iloc[2]['mode'] == 'fast_mobility'
def test_check_categories(self):
"""Asserts the correct identification of valid category dictionaries."""
tpls_file = os.path.join('tests', 'data', 'triplegs_transport_mode_identification.csv')
tpls = ti.read_triplegs_csv(tpls_file, sep=';', index_col="id")
correct_dict = {2: 'cat1', 7: 'cat2', np.inf: 'cat3'}
assert ti.analysis.transport_mode_identification.check_categories(correct_dict)
with pytest.raises(ValueError):
incorrect_dict = {10: 'cat1', 5: 'cat2', np.inf: 'cat3'}
tpls.as_triplegs.predict_transport_mode(method='simple-coarse', categories=incorrect_dict)
def test_trajectory_distance(self):
tpls = ti.read_triplegs_csv(
os.path.join('tests', 'data', 'geolife', 'geolife_triplegs.csv'))
D_single = calculate_distance_matrix(X=tpls.iloc[0:4],
dist_metric='dtw',
n_jobs=1)
D_multi = calculate_distance_matrix(X=tpls.iloc[0:4],
dist_metric='dtw',
n_jobs=4)
assert np.isclose(np.sum(np.abs(D_single - D_multi)), 0)
def test_check_categories(self):
"""Asserts the correct identification of valid category dictionaries."""
tpls_file = os.path.join("tests", "data",
"triplegs_transport_mode_identification.csv")
tpls = ti.read_triplegs_csv(tpls_file, sep=";", index_col="id")
correct_dict = {2: "cat1", 7: "cat2", np.inf: "cat3"}
assert _check_categories(correct_dict)
with pytest.raises(ValueError):
incorrect_dict = {10: "cat1", 5: "cat2", np.inf: "cat3"}
tpls.as_triplegs.predict_transport_mode(method="simple-coarse",
categories=incorrect_dict)
def test_trajectory_distance_via_accessor_xy(self):
tpls_file = os.path.join('tests', 'data', 'geolife',
'geolife_triplegs.csv')
tpls = ti.read_triplegs_csv(tpls_file, tz='utc', index_col='id')
x = tpls.iloc[0:2]
y = tpls.iloc[4:8]
D_single = x.as_triplegs.similarity(Y=y, dist_metric='dtw', n_jobs=1)
D_multi = x.as_triplegs.similarity(Y=y, dist_metric='dtw', n_jobs=4)
assert np.isclose(np.sum(np.abs(D_single - D_multi)), 0)
def test_simple_coarse_identification_wgs_84(self):
"""Asserts the correct behaviour with data in wgs84."""
tpls_file = os.path.join("tests", "data",
"triplegs_transport_mode_identification.csv")
tpls = ti.read_triplegs_csv(tpls_file, sep=";", index_col="id")
tpls_2 = tpls.set_crs(epsg=4326)
tpls_transport_mode_2 = tpls_2.as_triplegs.predict_transport_mode(
method="simple-coarse")
assert tpls_transport_mode_2.iloc[0]["mode"] == "slow_mobility"
assert tpls_transport_mode_2.iloc[1]["mode"] == "motorized_mobility"
assert tpls_transport_mode_2.iloc[2]["mode"] == "fast_mobility"
def test_triplegs_plot(self):
"""Use trackintel visualization function to plot triplegs and check if file exists."""
tmp_file = os.path.join('tests', 'data', 'triplegs_plot.png')
pfs_file = os.path.join('tests', 'data', 'positionfixes.csv')
pfs = ti.read_positionfixes_csv(pfs_file, sep=';', index_col='id', crs='EPSG:4326')
tpls_file = os.path.join('tests', 'data', 'triplegs.csv')
tpls = ti.read_triplegs_csv(tpls_file, sep=';', index_col='id', crs='EPSG:4326')
tpls.as_triplegs.plot(out_filename=tmp_file, positionfixes=pfs, plot_osm=False)
assert os.path.exists(tmp_file)
os.remove(tmp_file)
def test_triplegs_from_to_csv(self):
orig_file = 'tests/data/triplegs.csv'
tmp_file = 'tests/data/triplegs_test.csv'
tpls = ti.read_triplegs_csv(orig_file, sep=';')
tpls['started_at'] = tpls['started_at'].apply(
lambda d: d.isoformat().replace('+00:00', 'Z'))
tpls['finished_at'] = tpls['finished_at'].apply(
lambda d: d.isoformat().replace('+00:00', 'Z'))
tpls.as_triplegs.to_csv(
tmp_file,
sep=';',
columns=['user_id', 'started_at', 'finished_at', 'geom'])
assert filecmp.cmp(orig_file, tmp_file, shallow=False)
os.remove(tmp_file)
def test_smoothen_triplegs(self):
tpls_file = os.path.join("tests", "data", "triplegs_with_too_many_points_test.csv")
tpls = ti.read_triplegs_csv(tpls_file, sep=";", index_col=None)
tpls_smoothed = ti.preprocessing.triplegs.smoothen_triplegs(tpls, tolerance=0.0001)
line1 = tpls.iloc[0].geom
line1_smoothed = tpls_smoothed.iloc[0].geom
line2 = tpls.iloc[1].geom
line2_smoothed = tpls_smoothed.iloc[1].geom
assert line1.length == line1_smoothed.length
assert line2.length == line2_smoothed.length
assert len(line1.coords) == 10
assert len(line2.coords) == 7
assert len(line1_smoothed.coords) == 4
assert len(line2_smoothed.coords) == 3
def test_set_datatime_tz(self):
"""Test setting the timezone infomation when reading."""
# check if tz is added to the datatime column
file = os.path.join("tests", "data", "triplegs.csv")
tpls = ti.read_triplegs_csv(file, sep=";", index_col="id")
assert pd.api.types.is_datetime64tz_dtype(tpls["started_at"])
# check if a timezone will be set after manually deleting the timezone
tpls["started_at"] = tpls["started_at"].dt.tz_localize(None)
assert not pd.api.types.is_datetime64tz_dtype(tpls["started_at"])
tmp_file = os.path.join("tests", "data", "triplegs_test_2.csv")
tpls.as_triplegs.to_csv(tmp_file, sep=";")
tpls = ti.read_triplegs_csv(tmp_file,
sep=";",
index_col="id",
tz="utc")
assert pd.api.types.is_datetime64tz_dtype(tpls["started_at"])
# check if a warning is raised if 'tz' is not provided
with pytest.warns(UserWarning):
ti.read_triplegs_csv(tmp_file, sep=";", index_col="id")
os.remove(tmp_file)
def test_simple_coarse_identification_no_crs(self):
"""Assert that the simple-coarse transport mode identification throws the correct
warning and and yields the correct results for WGS84.
"""
tpls_file = os.path.join('tests', 'data', 'triplegs_transport_mode_identification.csv')
tpls = ti.read_triplegs_csv(tpls_file, sep=';', index_col="id")
with pytest.warns(UserWarning,
match='Your data is not projected. WGS84 is assumed and for length ' +
'calculation the haversine distance is used'):
tpls_transport_mode = tpls.as_triplegs.predict_transport_mode(method='simple-coarse')
assert tpls_transport_mode.iloc[0]['mode'] == 'slow_mobility'
assert tpls_transport_mode.iloc[1]['mode'] == 'motorized_mobility'
assert tpls_transport_mode.iloc[2]['mode'] == 'fast_mobility'
def test_read_triplegs_gpd(self):
"""Test if the results of reading from gpd and csv agrees."""
gdf = gpd.read_file(os.path.join("tests", "data", "triplegs.geojson"))
gdf.set_index("id", inplace=True)
tpls_from_gpd = ti.io.from_geopandas.read_triplegs_gpd(gdf,
user_id="User",
geom="geometry",
tz="utc")
tpls_file = os.path.join("tests", "data", "triplegs.csv")
tpls_from_csv = ti.read_triplegs_csv(tpls_file,
sep=";",
tz="utc",
index_col="id")
pd.testing.assert_frame_equal(tpls_from_gpd,
tpls_from_csv,
check_exact=False)
def test_simple_coarse_identification_no_crs(self):
"""
Assert that the simple-coarse transport mode identification throws the correct
warning and and yields the correct results for WGS84.
"""
tpls_file = os.path.join("tests", "data",
"triplegs_transport_mode_identification.csv")
tpls = ti.read_triplegs_csv(tpls_file, sep=";", index_col="id")
with pytest.warns(
UserWarning,
match="Your data is not projected.",
):
tpls_transport_mode = tpls.as_triplegs.predict_transport_mode(
method="simple-coarse")
assert tpls_transport_mode.iloc[0]["mode"] == "slow_mobility"
assert tpls_transport_mode.iloc[1]["mode"] == "motorized_mobility"
assert tpls_transport_mode.iloc[2]["mode"] == "fast_mobility"
