def test_clean_data_germany_goettingen_alt_is_fluke():
germany_data = c.fetch_data_germany(filter_goettingen_alt=False)
cleaned = c.fetch_data_germany(filter_goettingen_alt=True)
## could also use this command:
cleaned = c.clean_data_germany_remove_goettingen_alt(germany_data)
# how many rows did we delete?
n = len(germany_data) - len(cleaned)
if n == 1:
# Normal as of 167 Sept (see oscovida.cleane_data_germany_remove_göttingen_alt.__doc__)
pass
elif n > 1:
msg = f"We have found {n} rows of Göttingen alt data - please investigate\n" + \
"if this is a real / important LK in Germany"
raise ValueError(msg, germany_data, cleaned)
elif n == 0:
msg = "There are now rows with LK Göttingen (alt). \n" + \
"Consider removing the data cleaning code for Göttingen (alt)."
print(msg)
# should we raise an error here to notice this situation?
raise ValueError(msg)
else:
raise NotImplementedError("This should not be possible.", germany_data,
cleaned)
def generate_reports_germany(
*,
workers,
kernel_name,
wwwroot,
force,
disable_pbar,
debug,
incidence_period=7,
incidence_threshold=50,
):
_ = oscovida.fetch_data_germany()
# TODO: The get_x_list methods should be part of Reporter class
germany_regions = get_germany_regions_list()
# data cleaning: on 13 April, we had a Landkreis "LK Göttingen (alt)"
# with only one data point. This causes plots to fail, because there
# is nothing to plot, and then the legend() command failed.
# We assume that the RKI labels unusual data with '(alt)', and remove those.
alt_data_sets = ["(alt)" in r[1].lower() for r in germany_regions]
if any(alt_data_sets):
bad_datasets = list(compress(germany_regions, alt_data_sets))
logging.warning(
f"Removing datasets label with '(alt)': {bad_datasets}")
for bd in bad_datasets:
c, d = oscovida.germany_get_region(landkreis=bd[1])
logging.warning(
f"\tremoved: {bd} : len(cases)={len(c)}, len(deaths)={len(d)}")
bad_indices = list(compress(range(len(alt_data_sets)), alt_data_sets))
for i in sorted(bad_indices, reverse=True):
del germany_regions[i]
gre = ReportExecutor(
Reporter=GermanyReport,
kernel_name=kernel_name,
wwwroot=wwwroot,
expiry_hours=2,
attempts=3,
workers=workers,
force=force,
disable_pbar=disable_pbar,
debug=debug,
)
if debug:
germany_regions = germany_regions[:10]
gre.create_html_reports(germany_regions)
gre.create_markdown_index_page()
gre.create_markdown_incidence_page(period=incidence_period,
threshold=incidence_threshold)
def get_germany_subregion_list():
"""returns list of subregions (Kreise),
ordered according to (i) Land, then (ii) Kreis
"""
x = fetch_data_germany()
land_kreis = x[['Bundesland', 'Landkreis']]
ordered = land_kreis.sort_values(['Bundesland', 'Landkreis'])
return list(ordered['Landkreis'].drop_duplicates())
def test_germany_get_population():
germany = c.germany_get_population()
assert germany.index.name == 'county'
assert 'population' in germany.columns
assert 'cases7_per_100k' in germany.columns
germany_data = c.fetch_data_germany()
assert set(germany_data['Landkreis']) == set(germany.index)
hamburg = germany.loc['SK Hamburg'].population
assert hamburg > 1800000
pinneberg = germany.loc['LK Pinneberg'].population
assert pinneberg > 30000
def test_germany_get_population():
germany = c.germany_get_population()
assert germany.index.name == 'county'
assert 'population' in germany.columns
assert 'cases7_per_100k' in germany.columns
germany_data = c.fetch_data_germany()
assert set(germany_data['Landkreis']) == set(germany.index)
hamburg = germany.loc['SK Hamburg'].population
assert hamburg > 1800000
pinneberg = germany.loc['LK Pinneberg'].population
assert pinneberg > 30000
# https://github.com/oscovida/oscovida/issues/210
saarpfalz = germany.loc['LK Saarpfalz-Kreis'].population
assert saarpfalz > 130000
aachen = germany.loc['StadtRegion Aachen'].population
assert aachen > 500000
def germany_check_subregion__is_known(subregion):
d = oscovida.fetch_data_germany()
assert subregion in list(
d["Landkreis"].drop_duplicates()), f"{subregion} is unknown."
def germany_check_region_is_known(region):
d = oscovida.fetch_data_germany()
assert region in list(
d["Bundesland"].drop_duplicates()), f"{region} is unknown."
def get_germany_regions_list():
data_germany = oscovida.fetch_data_germany()
land_kreis = data_germany[["Bundesland", "Landkreis"]]
ordered = land_kreis.sort_values(["Bundesland", "Landkreis"])
return ordered.drop_duplicates().values.tolist()