def test_read_staypoints_csv_crs_parameter(self):
file = os.path.join('tests', 'data', 'staypoints.csv')
crs = "EPSG:2056"
stps = ti.read_staypoints_csv(file, sep=';', tz='utc', index_col="id")
assert stps.crs is None
stps = ti.read_staypoints_csv(file, sep=';', tz='utc', index_col="id", crs=crs)
assert stps.crs == crs
def test_staypoints_csv_index_col(self):
"""Test if `index_col` can be set."""
file = os.path.join('tests', 'data', 'staypoints.csv')
ind_name = 'id'
pfs = ti.read_staypoints_csv(file, sep=";", index_col=ind_name)
assert pfs.index.name == ind_name
pfs = ti.read_staypoints_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", "staypoints.csv")
ind_name = "id"
pfs = ti.read_staypoints_csv(file, sep=";", index_col=ind_name)
assert pfs.index.name == ind_name
pfs = ti.read_staypoints_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", "staypoints.csv")
crs = "EPSG:2056"
sp = ti.read_staypoints_csv(file, sep=";", tz="utc", index_col="id")
assert sp.crs is None
sp = ti.read_staypoints_csv(file,
sep=";",
tz="utc",
index_col="id",
crs=crs)
assert sp.crs == crs
def test_staypoints_from_to_csv(self):
orig_file = os.path.join('tests', 'data', 'staypoints.csv')
mod_file = os.path.join('tests', 'data', 'staypoints_mod_columns.csv')
tmp_file = os.path.join('tests', 'data', 'staypoints_test.csv')
stps = ti.read_staypoints_csv(orig_file, sep=';', tz='utc', index_col="id")
mod_stps = ti.read_staypoints_csv(mod_file, columns={'User': '******'}, sep=';', index_col="id")
assert mod_stps.equals(stps)
stps['started_at'] = stps['started_at'].apply(lambda d: d.isoformat().replace('+00:00', 'Z'))
stps['finished_at'] = stps['finished_at'].apply(lambda d: d.isoformat().replace('+00:00', 'Z'))
columns = ['user_id', 'started_at', 'finished_at', 'elevation', 'geom']
stps.as_staypoints.to_csv(tmp_file, sep=';', columns=columns)
assert filecmp.cmp(orig_file, tmp_file, shallow=False)
os.remove(tmp_file)
def test_generate_locations_dtype_consistent(self):
"""Test the dtypes for the generated columns."""
stps_file = os.path.join("tests", "data", "geolife",
"geolife_staypoints.csv")
stps = ti.read_staypoints_csv(stps_file, tz="utc", index_col="id")
#
stps, locs = stps.as_staypoints.generate_locations(
method="dbscan",
epsilon=10,
num_samples=0,
distance_metric="haversine",
agg_level="dataset")
assert stps["user_id"].dtype == locs["user_id"].dtype
assert stps["location_id"].dtype == "Int64"
assert locs.index.dtype == "int64"
# change the user_id to string
stps["user_id"] = stps["user_id"].apply(lambda x: str(x))
stps, locs = stps.as_staypoints.generate_locations(
method="dbscan",
epsilon=10,
num_samples=0,
distance_metric="haversine",
agg_level="dataset")
assert stps["user_id"].dtype == locs["user_id"].dtype
assert stps["location_id"].dtype == "Int64"
assert locs.index.dtype == "int64"
def test_generate_locations_dbscan_hav_euc(self):
stps_file = os.path.join("tests", "data", "geolife",
"geolife_staypoints.csv")
stps = ti.read_staypoints_csv(stps_file, tz="utc", index_col="id")
# haversine calculation
_, loc_har = stps.as_staypoints.generate_locations(
method="dbscan",
epsilon=100,
num_samples=0,
distance_metric="haversine",
agg_level="dataset")
# WGS_1984
stps.crs = "epsg:4326"
# WGS_1984_UTM_Zone_49N
stps = stps.to_crs("epsg:32649")
# euclidean calculation
_, loc_eu = stps.as_staypoints.generate_locations(
method="dbscan",
epsilon=100,
num_samples=0,
distance_metric="euclidean",
agg_level="dataset")
assert len(loc_har) == len(
loc_eu
), "The #location should be the same for haversine" + "and euclidean distances"
def test_filter_staypoints(self):
# read staypoints and area file
spts_file = os.path.join('tests', 'data', 'geolife',
'geolife_staypoints.csv')
spts = ti.read_staypoints_csv(spts_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
spts.crs = 'epsg:4326'
within_spts = spts.as_staypoints.spatial_filter(areas=extent,
method="within",
re_project=True)
intersects_spts = spts.as_staypoints.spatial_filter(
areas=extent, method="intersects", re_project=True)
crosses_spts = spts.as_staypoints.spatial_filter(areas=extent,
method="crosses",
re_project=True)
# the result obtained from ArcGIS
gis_within_num = 13
assert len(within_spts) == gis_within_num, "The spatial filtered sp number should be the same as" + \
"the one from the result with ArcGIS"
assert all(within_spts.geometry == intersects_spts.geometry), "For sp the result of within and" + \
"intersects should be the same"
assert len(crosses_spts) == 0, "There will be no point crossing area"
def test_cluster_staypoints_dbscan_user_dataset(self):
spts = ti.read_staypoints_csv(
'tests/data/geolife/geolife_staypoints.csv')
# take the first row and duplicate once
spts = spts.head(1)
spts = spts.append(spts, ignore_index=True)
# assign a different user_id to the second row
spts.iloc[1, 5] = 1
# duplicate for a certain number
spts = spts.append([spts] * 5, ignore_index=True)
_, locs_ds = spts.as_staypoints.extract_locations(
method='dbscan',
epsilon=10,
num_samples=0,
distance_matrix_metric='haversine',
agg_level='dataset')
_, locs_us = spts.as_staypoints.extract_locations(
method='dbscan',
epsilon=10,
num_samples=0,
distance_matrix_metric='haversine',
agg_level='user')
loc_ds_num = locs_ds['location_id'].unique().shape[0]
loc_us_num = locs_us['location_id'].unique().shape[0]
assert loc_ds_num == 1, "Considering all staypoints at once, there should be only one location"
assert loc_us_num == 2, "Considering user staypoints separately, there should be two locations"
def test_generate_locations_dtype_consistent(self):
"""Test the dtypes for the generated columns."""
stps_file = os.path.join('tests', 'data', 'geolife',
'geolife_staypoints.csv')
stps = ti.read_staypoints_csv(stps_file, tz='utc', index_col='id')
#
stps, locs = stps.as_staypoints.generate_locations(
method='dbscan',
epsilon=10,
num_samples=0,
distance_matrix_metric='haversine',
agg_level='dataset')
assert stps['user_id'].dtype == locs['user_id'].dtype
assert stps['location_id'].dtype == 'Int64'
assert locs.index.dtype == 'int64'
# change the user_id to string
stps['user_id'] = stps['user_id'].apply(lambda x: str(x))
stps, locs = stps.as_staypoints.generate_locations(
method='dbscan',
epsilon=10,
num_samples=0,
distance_matrix_metric='haversine',
agg_level='dataset')
assert stps['user_id'].dtype == locs['user_id'].dtype
assert stps['location_id'].dtype == 'Int64'
assert locs.index.dtype == 'int64'
def test_generate_locations_dbscan_user_dataset(self):
stps_file = os.path.join('tests', 'data', 'geolife',
'geolife_staypoints.csv')
stps = ti.read_staypoints_csv(stps_file, tz='utc', index_col='id')
# take the first row and duplicate once
stps = stps.head(1)
stps = stps.append(stps, ignore_index=True)
# assign a different user_id to the second row
stps.iloc[1, 4] = 1
# duplicate for a certain number
stps = stps.append([stps] * 5, ignore_index=True)
_, locs_ds = stps.as_staypoints.generate_locations(
method='dbscan',
epsilon=10,
num_samples=0,
distance_matrix_metric='haversine',
agg_level='dataset')
_, locs_us = stps.as_staypoints.generate_locations(
method='dbscan',
epsilon=10,
num_samples=0,
distance_matrix_metric='haversine',
agg_level='user')
loc_ds_num = len(locs_ds.index.unique())
loc_us_num = len(locs_us.index.unique())
assert loc_ds_num == 1, "Considering all staypoints at once, there should be only one location"
assert loc_us_num == 2, "Considering user staypoints separately, there should be two locations"
def test_dbscan_user_dataset(self):
"""Test user and dataset location generation."""
sp_file = os.path.join("tests", "data", "geolife",
"geolife_staypoints.csv")
sp = ti.read_staypoints_csv(sp_file, tz="utc", index_col="id")
# take the first row and duplicate once
sp = sp.head(1)
sp = sp.append(sp, ignore_index=True)
# assign a different user_id to the second row
sp.iloc[1, 4] = 1
# duplicate for a certain number
sp = sp.append([sp] * 5, ignore_index=True)
_, locs_ds = sp.as_staypoints.generate_locations(
method="dbscan",
epsilon=10,
num_samples=0,
distance_metric="haversine",
agg_level="dataset")
_, locs_us = sp.as_staypoints.generate_locations(
method="dbscan",
epsilon=10,
num_samples=0,
distance_metric="haversine",
agg_level="user")
loc_dataset_num = len(locs_ds.index.unique())
loc_user_num = len(locs_us.index.unique())
assert loc_dataset_num == 1
assert loc_user_num == 2
def test_dbscan_hav_euc(self):
"""Test if using haversine and euclidean distances will generate the same location result."""
sp_file = os.path.join("tests", "data", "geolife",
"geolife_staypoints.csv")
sp = ti.read_staypoints_csv(sp_file, tz="utc", index_col="id")
# haversine calculation
_, loc_har = sp.as_staypoints.generate_locations(
method="dbscan",
epsilon=100,
num_samples=0,
distance_metric="haversine",
agg_level="dataset")
# WGS_1984
sp.crs = "epsg:4326"
# WGS_1984_UTM_Zone_49N
sp = sp.to_crs("epsg:32649")
# euclidean calculation
_, loc_eu = sp.as_staypoints.generate_locations(
method="dbscan",
epsilon=100,
num_samples=0,
distance_metric="euclidean",
agg_level="dataset")
assert len(loc_har) == len(loc_eu)
def test_keyword_combinations(self):
spts_file = os.path.join('tests', 'data', 'geolife',
'geolife_staypoints.csv')
spts = ti.read_staypoints_csv(spts_file, tz='utc', index_col='id')
x = spts.iloc[0:5]
y = spts.iloc[5:15]
_ = calculate_distance_matrix(X=x,
Y=y,
dist_metric='euclidean',
n_jobs=-1)
_ = calculate_distance_matrix(X=y,
Y=x,
dist_metric='haversine',
n_jobs=-1)
d_mink1 = calculate_distance_matrix(X=x,
Y=x,
dist_metric='minkowski',
p=1)
d_mink2 = calculate_distance_matrix(X=x,
Y=x,
dist_metric='minkowski',
p=2)
d_euc = calculate_distance_matrix(X=x, Y=x, dist_metric='euclidean')
assert not np.array_equal(d_mink1, d_mink2)
assert np.array_equal(d_euc, d_mink2)
def test_cluster_staypoints_dbscan_loc(self):
spts = ti.read_staypoints_csv(
'tests/data/geolife/geolife_staypoints.csv')
spts, locs = spts.as_staypoints.extract_locations(
method='dbscan',
epsilon=10,
num_samples=0,
distance_matrix_metric='haversine',
agg_level='dataset')
# create locations as grouped staypoints, another way to create locations
other_locs = pd.DataFrame(columns=['user_id', 'location_id', 'center'])
grouped_df = spts.groupby(['user_id', 'location_id'])
for combined_id, group in grouped_df:
user_id, location_id = combined_id
group.set_geometry(spts.geometry.name, inplace=True)
if int(location_id) != -1:
temp_loc = {}
temp_loc['user_id'] = user_id
temp_loc['location_id'] = location_id
# point geometry of place
temp_loc['center'] = Point(group.geometry.x.mean(),
group.geometry.y.mean())
other_locs = other_locs.append(temp_loc, ignore_index=True)
other_locs = gpd.GeoDataFrame(other_locs,
geometry='center',
crs=spts.crs)
assert all(other_locs['center'] ==
locs['center']), "The location geometry should be the same"
assert all(other_locs['location_id'] == locs['location_id']), "The location id should be the same" + \
"and start from one"
def test_keyword_combinations(self):
stps_file = os.path.join("tests", "data", "geolife",
"geolife_staypoints.csv")
stps = ti.read_staypoints_csv(stps_file, tz="utc", index_col="id")
x = stps.iloc[0:5]
y = stps.iloc[5:15]
_ = calculate_distance_matrix(X=x,
Y=y,
dist_metric="euclidean",
n_jobs=-1)
_ = calculate_distance_matrix(X=y,
Y=x,
dist_metric="haversine",
n_jobs=-1)
d_mink1 = calculate_distance_matrix(X=x,
Y=x,
dist_metric="minkowski",
p=1)
d_mink2 = calculate_distance_matrix(X=x,
Y=x,
dist_metric="minkowski",
p=2)
d_euc = calculate_distance_matrix(X=x, Y=x, dist_metric="euclidean")
assert not np.array_equal(d_mink1, d_mink2)
assert np.array_equal(d_euc, d_mink2)
def test_filter_staypoints(self):
"""Test if spatial_filter works for staypoints."""
# read staypoints and area file
sp_file = os.path.join("tests", "data", "geolife",
"geolife_staypoints.csv")
sp = ti.read_staypoints_csv(sp_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
sp.crs = "epsg:4326"
within_sp = sp.as_staypoints.spatial_filter(areas=extent,
method="within",
re_project=True)
intersects_sp = sp.as_staypoints.spatial_filter(areas=extent,
method="intersects",
re_project=True)
crosses_sp = sp.as_staypoints.spatial_filter(areas=extent,
method="crosses",
re_project=True)
# the result obtained from ArcGIS
gis_within_num = 13
assert len(within_sp) == gis_within_num, (
"The spatial filtered sp number should be the same as" +
"the one from the result with ArcGIS")
assert len(crosses_sp) == 0, "There will be no point crossing area"
# For staypoints the result of within and intersects should be the same
assert_geodataframe_equal(within_sp,
intersects_sp,
check_less_precise=True)
def test_generate_locations_dbscan_hav_euc(self):
stps_file = os.path.join('tests', 'data', 'geolife',
'geolife_staypoints.csv')
stps = ti.read_staypoints_csv(stps_file, tz='utc', index_col='id')
# haversine calculation
_, loc_har = stps.as_staypoints.generate_locations(
method='dbscan',
epsilon=100,
num_samples=0,
distance_matrix_metric='haversine',
agg_level='dataset')
# WGS_1984
stps.crs = 'epsg:4326'
# WGS_1984_UTM_Zone_49N
stps = stps.to_crs("epsg:32649")
# euclidean calculation
_, loc_eu = stps.as_staypoints.generate_locations(
method='dbscan',
epsilon=100,
num_samples=0,
distance_matrix_metric='euclidean',
agg_level='dataset')
assert len(loc_har) == len(loc_eu) , "The #location should be the same for haversine" + \
"and euclidean distances"
def test_dbscan_loc(self):
"""Test haversine dbscan location result with manually grouping the locations method."""
stps_file = os.path.join("tests", "data", "geolife", "geolife_staypoints.csv")
stps = ti.read_staypoints_csv(stps_file, tz="utc", index_col="id")
stps, locs = stps.as_staypoints.generate_locations(
method="dbscan", epsilon=10, num_samples=0, distance_metric="haversine", agg_level="dataset"
)
# create locations as grouped staypoints, another way to create locations
other_locs = pd.DataFrame(columns=["user_id", "id", "center"])
grouped_df = stps.groupby(["user_id", "location_id"])
for combined_id, group in grouped_df:
user_id, location_id = combined_id
group.set_geometry(stps.geometry.name, inplace=True)
if int(location_id) != -1:
temp_loc = {}
temp_loc["user_id"] = user_id
temp_loc["id"] = location_id
# point geometry of place
temp_loc["center"] = Point(group.geometry.x.mean(), group.geometry.y.mean())
other_locs = other_locs.append(temp_loc, ignore_index=True)
other_locs = gpd.GeoDataFrame(other_locs, geometry="center", crs=stps.crs)
other_locs.set_index("id", inplace=True)
assert all(other_locs["center"] == locs["center"])
assert all(other_locs.index == locs.index)
def test_haversine_vectorized(self):
spts = ti.read_staypoints_csv(
os.path.join('tests', 'data', 'geolife', 'geolife_staypoints.csv'))
x = spts.geometry.x.values
y = spts.geometry.y.values
n = len(x)
# our distance
ix_1, ix_2 = np.triu_indices(n, k=1)
x1 = x[ix_1]
y1 = y[ix_1]
x2 = x[ix_2]
y2 = y[ix_2]
d_ours = haversine_dist(x1, y1, x2, y2)
# their distance
x_rad = np.asarray([radians(_) for _ in x])
y_rad = np.asarray([radians(_) for _ in y])
yx = np.concatenate((y_rad.reshape(-1, 1), x_rad.reshape(-1, 1)),
axis=1)
D_theirs = haversine_distances(yx, yx) * 6371000
d_theirs = D_theirs[ix_1, ix_2]
assert np.sum(
np.abs(d_ours -
d_theirs)) < 0.01 # 1cm for 58 should be good enough
def test_cluster_staypoints_dbscan_hav_euc(self):
spts = ti.read_staypoints_csv(
os.path.join('tests', 'data', 'geolife', 'geolife_staypoints.csv'))
# haversine calculation
_, loc_har = spts.as_staypoints.extract_locations(
method='dbscan',
epsilon=100,
num_samples=0,
distance_matrix_metric='haversine',
agg_level='dataset')
# WGS_1984
spts.crs = 'epsg:4326'
# WGS_1984_UTM_Zone_49N
spts = spts.to_crs("epsg:32649")
# euclidean calculation
_, loc_eu = spts.as_staypoints.extract_locations(
method='dbscan',
epsilon=100,
num_samples=0,
distance_matrix_metric='euclidean',
agg_level='dataset')
assert len(loc_har) == len(loc_eu) , "The #location should be the same for haversine" + \
"and euclidean distances"
def test_as_staypoints_accessor(self):
orig_file = 'tests/data/staypoints.csv'
stps = ti.read_staypoints_csv(orig_file, sep=';')
assert stps.as_staypoints
stps = stps.drop(['geom'], axis=1)
with pytest.raises(AttributeError):
stps.as_staypoints
def test_as_staypoints_accessor(self):
stps_file = os.path.join('tests', 'data', 'staypoints.csv')
stps = ti.read_staypoints_csv(stps_file, sep=';', index_col='id')
assert stps.as_staypoints
stps = stps.drop(['geom'], axis=1)
with pytest.raises(AttributeError):
stps.as_staypoints
def test_staypoints_from_gpd(self):
gdf = gpd.read_file(os.path.join('tests', 'data', 'staypoints.geojson'))
gdf.set_index('id', inplace=True)
stps_from_gpd = ti.io.from_geopandas.staypoints_from_gpd(gdf, 'start_time', 'end_time', geom='geometry',
tz='utc')
stps_file = os.path.join('tests', 'data', 'staypoints.csv')
stps_from_csv = ti.read_staypoints_csv(stps_file, sep=';', tz='utc', index_col='id')
pd.testing.assert_frame_equal(stps_from_gpd, stps_from_csv, check_exact=False)
def test_create_activity_flag(self):
spts_test = ti.read_staypoints_csv(
os.path.join('tests', 'data', 'geolife', 'geolife_staypoints.csv'))
activity_true = spts_test['activity'].copy()
spts_test['activity'] = False
spts_test = spts_test.as_staypoints.create_activity_flag()
pd.testing.assert_series_equal(spts_test['activity'], activity_true)
def testdata_locs():
"""Read location test data from files."""
sp_file = os.path.join("tests", "data", "geolife",
"geolife_staypoints.csv")
sp = ti.read_staypoints_csv(sp_file, tz="utc", index_col="id")
sp, locs = sp.as_staypoints.generate_locations(method="dbscan",
epsilon=10,
num_samples=0,
distance_metric="haversine",
agg_level="dataset")
return locs
def test_staypoints_plot(self):
tmp_file = 'tests/data/staypoints_plot.png'
pfs = ti.read_positionfixes_csv('tests/data/positionfixes.csv',
sep=';')
stps = ti.read_staypoints_csv('tests/data/staypoints.csv', sep=';')
stps.as_staypoints.plot(out_filename=tmp_file,
radius=0.01,
positionfixes=pfs,
plot_osm=False)
assert os.path.exists(tmp_file)
os.remove(tmp_file)
def test_create_activity_flag(self):
"""Test if 'activity' = True is assigned to staypoints."""
stps_file = os.path.join("tests", "data", "geolife",
"geolife_staypoints.csv")
stps_test = ti.read_staypoints_csv(stps_file, tz="utc", index_col="id")
activity_true = stps_test["activity"].copy()
stps_test["activity"] = False
stps_test = stps_test.as_staypoints.create_activity_flag()
pd.testing.assert_series_equal(stps_test["activity"], activity_true)
def test_print_progress_flag(self, capsys):
"""Test if the print_progress bar controls the printing behavior."""
file = os.path.join("tests", "data", "geolife", "geolife_staypoints.csv")
staypoints = ti.read_staypoints_csv(file, tz="utc", index_col="id")
staypoints.as_staypoints.generate_locations(print_progress=True)
captured_print = capsys.readouterr()
assert captured_print.err != ""
staypoints.as_staypoints.generate_locations(print_progress=False)
captured_print = capsys.readouterr()
assert captured_print.err == ""
def test_staypoints_plot(self):
"""Use trackintel visualization function to plot staypoints and check if file exists."""
tmp_file = os.path.join('tests', 'data', 'staypoints_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')
stps_file = os.path.join('tests', 'data', 'staypoints.csv')
stps = ti.read_staypoints_csv(stps_file, sep=';', index_col='id', crs='EPSG:4326')
stps.as_staypoints.plot(out_filename=tmp_file, radius=0.01, positionfixes=pfs, plot_osm=False)
assert os.path.exists(tmp_file)
os.remove(tmp_file)