class TestGenericSeriesData_RateGroup_TS(unittest.TestCase):
def setUp(self):
self.estat = Request()
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_rg/generic/ecb_exr_rg_ts.xml')
self.resp = self.estat.get(fromfile=filepath)
def test_groups(self):
data = self.resp.data
self.assertEqual(len(list(data.groups)), 5)
self.assertEqual(len(list(data.series)), 4)
g2 = list(data.groups)[2]
self.assertEqual(g2.key.CURRENCY, 'GBP')
self.assertEqual(
g2.attrib.TITLE, 'ECB reference exchange rate, U.K. Pound sterling /Euro')
# Check group attributes of a series
s = list(data.series)[0]
g_attrib = s.group_attrib
self.assertEqual(len(g_attrib), 5)
self.assertIsInstance(g_attrib, tuple)
self.assertEqual(len(g_attrib), 5)
def test_footer(self):
filepath = os.path.join(
pkg_path, 'data/estat/footer.xml')
resp = self.estat.get(
fromfile=filepath, get_footer_url=None)
f = resp.footer
assert f.code == 413
assert f.severity == 'Infomation'
assert f.text[1].startswith('http')
class TestStructSpecSeriesData_RateGroup_TS(unittest.TestCase):
def setUp(self):
self.estat = Request()
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_rg/structured/ecb_exr_rg_ts.xml')
dsd_resp = self.estat.datastructure(
fromfile=os.path.join(pkg_path, 'data/exr/ecb_exr_rg/ecb_exr_rg.xml'))
dsd = dsd_resp.datastructure.DataStructure
self.resp = self.estat.get(fromfile=filepath, dsd=dsd)
def test_groups(self):
data = self.resp.data
self.assertEqual(len(list(data.groups)), 5)
self.assertEqual(len(list(data.series)), 4)
g2 = list(data.groups)[2]
self.assertEqual(g2.key.CURRENCY, 'GBP')
self.assertEqual(
g2.attrib.TITLE, 'ECB reference exchange rate, U.K. Pound sterling /Euro')
# Check group attributes of a series
s = list(data.series)[0]
g_attrib = s.group_attrib
self.assertEqual(len(g_attrib), 5)
self.assertIsInstance(g_attrib, tuple)
self.assertEqual(len(g_attrib), 5)
def test_samples_request(self):
"""Test the samples from the SDMXJSON spec."""
req = Request()
for name, data in sample_data.items():
resp = req.get(fromfile=self._filepath(name))
df = resp.write()
assert df.equals(data), \
'\n'.join(map(str, [name, df.index, data.index,
getattr(df, 'columns', ''),
getattr(data, 'columns', '')]))
class Test_ESTAT_dsd_apro_mk_cola(unittest.TestCase):
def setUp(self):
self.estat = Request('ESTAT')
filepath = os.path.join(test_path, 'data/estat/apro_dsd.xml')
self.resp = self.estat.get(fromfile=filepath)
def test_codelists_keys(self):
self.assertEqual(len(self.resp.codelist), 6)
self.assertIsInstance(self.resp.codelist.CL_GEO, model.Codelist)
def test_codelist_name(self):
self.assertEqual(
self.resp.msg.codelist.CL_GEO.UK.name.en, 'United Kingdom')
assert self.resp.codelist.CL_FREQ.name.en == 'FREQ'
def test_code_cls(self):
self.assertIsInstance(
self.resp.codelist.CL_FREQ.D, model.Code)
def test_writer(self):
df = self.resp.write(rows='codelist')
self.assertEqual(df.shape, (79, 2))
def tearDown(self):
pass
class test_dsd_common(unittest.TestCase):
def setUp(self):
self.estat = Request('ESTAT')
filepath = os.path.join(test_path, 'data/common/common.xml')
self.resp = self.estat.get(fromfile=filepath)
def test_codelists_keys(self):
self.assertEqual(len(self.resp.msg.codelist), 5)
self.assertIsInstance(self.resp.msg.codelist.CL_FREQ, model.Codelist)
def test_codelist_name(self):
self.assertEqual(self.resp.msg.codelist.CL_FREQ.D.name.en, 'Daily')
def test_code_cls(self):
self.assertIsInstance(self.resp.msg.codelist.CL_FREQ.D, model.Code)
def test_annotations(self):
code = self.resp.codelist.CL_FREQ.A
anno_list = list(code.annotations)
self.assertEqual(len(anno_list), 1)
a = anno_list[0]
self.assertIsInstance(a, model.Annotation)
self.assertTrue(a.text.en.startswith('It is'))
self.assertEqual(a.annotationtype, 'NOTE')
def setUp(self):
self.estat = Request()
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_rg/structured/ecb_exr_rg_ts.xml')
dsd_resp = self.estat.datastructure(
fromfile=os.path.join(pkg_path, 'data/exr/ecb_exr_rg/ecb_exr_rg.xml'))
dsd = dsd_resp.datastructure.DataStructure
self.resp = self.estat.get(fromfile=filepath, dsd=dsd)
def setUp(self):
self.estat = Request('ESTAT')
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_ng/structured/ecb_exr_ng_ts_gf.xml')
dsd_resp = self.estat.datastructure(
fromfile=os.path.join(pkg_path, 'data/exr/ecb_exr_ng/ecb_exr_ng_full.xml'))
dsd = dsd_resp.datastructure.DataStructure
self.resp = self.estat.data(fromfile=filepath, dsd=dsd)
class TestStructSpecSeriesDataSet2(unittest.TestCase):
def setUp(self):
self.estat = Request('ESTAT')
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_ng/structured/ecb_exr_ng_ts.xml')
dsd_resp = self.estat.datastructure(
fromfile=os.path.join(pkg_path, 'data/exr/ecb_exr_ng/ecb_exr_ng_full.xml'))
dsd = dsd_resp.datastructure.DataStructure
self.resp = self.estat.data(fromfile=filepath, dsd=dsd)
def test_header_attributes(self):
self.assertEqual(self.resp.header.structured_by, 'STR1')
self.assertEqual(self.resp.header.dim_at_obs, 'TIME_PERIOD')
def test_dataset_cls(self):
self.assertIsInstance(self.resp.data, model.DataSet)
def test_structured_obs(self):
data = self.resp.data
# empty obs iterator
self.assertEqual(len(list(data.obs())), 0)
series_list = list(data.series)
self.assertEqual(len(series_list), 4)
s3 = series_list[3]
self.assertIsInstance(s3, model.Series)
self.assertIsInstance(s3.key, tuple)
self.assertEqual(len(s3.key), 4)
self.assertEqual(s3.key.CURRENCY, 'USD')
self.assertEqual(s3.attrib.DECIMALS, '4')
obs_list = list(s3.obs(reverse_obs=True))
self.assertEqual(len(obs_list), 3)
o0 = obs_list[2]
self.assertEqual(len(o0), 3)
self.assertEqual(o0.dim, '2010-08')
self.assertEqual(o0.value, '1.2894')
self.assertIsInstance(o0.attrib, tuple)
self.assertEqual(o0.attrib.OBS_STATUS, 'A')
def test_dataframe(self):
data = self.resp.data
df = self.resp.write(
data, attributes='', asframe=True, reverse_obs=True)
self.assertIsInstance(df, pandas.core.frame.DataFrame)
self.assertEqual(df.shape, (3, 4))
def test_write_source(self):
"""Test the write_source() method."""
req = Request()
for name in sample_data.keys():
orig_fn = self._filepath(name)
temp_fn = self._filepath(name + '-write-source')
# Read the message
resp = req.get(fromfile=orig_fn)
# Write to a temporary JSON file
resp.write_source(temp_fn)
# Read the two files and compare JSON (ignores ordering)
with open(orig_fn) as orig, open(temp_fn) as temp:
assert json.load(orig) == json.load(temp)
# Delete the temporary file
os.remove(temp_fn)
class TestStructSpecFlatDataSet(unittest.TestCase):
def setUp(self):
self.estat = Request('ESTAT')
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_ng/structured/ecb_exr_ng_flat.xml')
dsd_resp = self.estat.datastructure(
fromfile=os.path.join(pkg_path, 'data/exr/ecb_exr_ng/ecb_exr_ng_full.xml'))
dsd = dsd_resp.datastructure.DataStructure
self.resp = self.estat.get(fromfile=filepath, dsd=dsd)
def test_msg_type(self):
self.assertIsInstance(self.resp.msg, model.DataMessage)
def test_header_attributes(self):
self.assertEqual(self.resp.header.structured_by, 'STR1')
self.assertEqual(self.resp.header.dim_at_obs, 'AllDimensions')
def test_dataset_cls(self):
self.assertIsInstance(self.resp.data, model.DataSet)
self.assertEqual(self.resp.msg.data.dim_at_obs, 'AllDimensions')
def test_generic_obs(self):
data = self.resp.data
# empty series list
self.assertEqual(len(list(data.series)), 0)
obs_list = list(data.obs())
self.assertEqual(len(obs_list), 12)
o0 = obs_list[0]
self.assertEqual(len(o0), 3)
self.assertIsInstance(o0.key, tuple) # obs_key
self.assertEqual(o0.key.FREQ, 'M')
self.assertEqual(o0.key.CURRENCY, 'CHF')
self.assertEqual(o0.value, '1.3413')
self.assertIsInstance(o0.attrib, tuple)
self.assertEqual(o0.attrib.OBS_STATUS, 'A')
self.assertEqual(o0.attrib.DECIMALS, '4')
def test_write2pandas(self):
data_series = self.resp.write(attributes='',
asframe=False, reverse_obs=False)
self.assertIsInstance(data_series, pandas.Series)
class TestGenericSeriesData_SiblingGroup_TS(unittest.TestCase):
def setUp(self):
self.estat = Request()
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_sg/generic/ecb_exr_sg_ts.xml')
self.resp = self.estat.get(fromfile=filepath)
def test_groups(self):
data = self.resp.data
self.assertEqual(len(list(data.groups)), 4)
self.assertEqual(len(list(data.series)), 4)
g2 = list(data.groups)[2]
self.assertEqual(g2.key.CURRENCY, 'JPY')
self.assertEqual(
g2.attrib.TITLE, 'ECB reference exchange rate, Japanese yen/Euro')
# Check group attributes of a series
s = list(data.series)[0]
g_attrib = s.group_attrib
self.assertEqual(len(g_attrib), 1)
self.assertIsInstance(g_attrib, tuple)
self.assertEqual(len(g_attrib), 1)
class test_exr_constraints(unittest.TestCase):
def setUp(self):
self.ecb = Request('ecb')
filepath = os.path.join(test_path, 'data/exr_flow.xml')
self.resp = self.ecb.get(fromfile=filepath)
def test_constrained_codes(self):
m = self.resp.msg
self.assertEqual(m._dim_ids[0], 'FREQ')
self.assertEqual(len(m._dim_ids), 5)
self.assertEqual(len(m._dim_ids), 5)
self.assertEqual(len(m._dim_codes), 5)
self.assertEqual(len(m._attr_ids), 9)
self.assertEqual(len(m._attr_codes), 9)
self.assertEqual(m._attr_ids[-1], 'UNIT_MULT')
self.assertIn('5', m._attr_codes.UNIT_MULT)
self.assertIn('W', m._dim_codes.FREQ)
self.assertIn('W', m._dim_codes.FREQ)
self.assertEqual(len(m._constrained_codes), 14)
self.assertNotIn('W', m._constrained_codes.FREQ)
key = {'FREQ': ['W']}
self.assertTrue(m.in_codes(key))
self.assertFalse(m.in_constraints(key, raise_error=False))
self.assertRaises(ValueError, m.in_constraints, key)
self.assertTrue(m.in_constraints({'CURRENCY': ['CHF']}))
# test with invalid key
self.assertRaises(TypeError, m._in_constraints, {'FREQ': 'A'})
# structure writer with constraints
out = self.resp.write()
cl = out.codelist
self.assertEqual(cl.shape, (3555, 2))
# unconstrained codelists
out = self.resp.write(constraint=False)
cl = out.codelist
self.assertEqual(cl.shape, (4177, 2))
def setUp(self):
self.ecb = Request('ecb')
filepath = os.path.join(test_path, 'data/exr_flow.xml')
self.resp = self.ecb.get(fromfile=filepath)
class InseeTestCase(unittest.TestCase):
# nosetests -s -v pandasdmx.tests.test_insee:InseeTestCase
def setUp(self):
unittest.TestCase.setUp(self)
self.sdmx = Request('INSEE')
def test_load_dataset(self):
dataset_code = 'IPI-2010-A21'
'''load all dataflows'''
dataflows_response = self.sdmx.get(
resource_type='dataflow', agency='FR1', fromfile=DATAFLOW_FP)
dataflows = dataflows_response.msg.dataflow
self.assertEqual(len(dataflows.keys()), 663)
self.assertTrue(dataset_code in dataflows)
'''load datastructure for current dataset_code'''
fp_datastructure = DATASETS[dataset_code]['datastructure-fp']
datastructure_response = self.sdmx.get(
resource_type='datastructure', agency='FR1', fromfile=fp_datastructure)
self.assertTrue(
dataset_code in datastructure_response.msg.datastructure)
dsd = datastructure_response.msg.datastructure[dataset_code]
'''Verify dimensions list'''
dimensions = OrderedDict([dim.id, dim] for dim in dsd.dimensions.aslist(
) if dim.id not in ['TIME', 'TIME_PERIOD'])
dim_keys = list(dimensions.keys())
self.assertEqual(dim_keys, ['FREQ', 'PRODUIT', 'NATURE'])
'''load datas for the current dataset'''
fp_data = DATASETS[dataset_code]['data-fp']
data = self.sdmx.get(
resource_type='data', agency='FR1', fromfile=fp_data)
'''Verify series count and values'''
series = list(data.msg.data.series)
series_count = len(series)
self.assertEqual(series_count, DATASETS[dataset_code]['series_count'])
first_series = series[0]
observations = list(first_series.obs())
first_obs = observations[0]
last_obs = observations[-1]
self.assertEqual(first_obs.dim, '2015-10')
self.assertEqual(first_obs.value, '105.61')
self.assertEqual(last_obs.dim, '1990-01')
self.assertEqual(last_obs.value, '139.22')
def test_fixe_key_names(self):
"""Verify key or attribute contains '-' in name
"""
dataset_code = 'CNA-2010-CONSO-SI-A17'
fp_datastructure = DATASETS[dataset_code]['datastructure-fp']
datastructure_response = self.sdmx.get(
resource_type='datastructure', agency='FR1', fromfile=fp_datastructure)
self.assertTrue(
dataset_code in datastructure_response.msg.datastructure)
dsd = datastructure_response.msg.datastructure[dataset_code]
dimensions = OrderedDict([dim.id, dim] for dim in dsd.dimensions.aslist(
) if dim.id not in ['TIME', 'TIME_PERIOD'])
dim_keys = list(dimensions.keys())
self.assertEqual(
dim_keys, ['SECT-INST', 'OPERATION', 'PRODUIT', 'PRIX'])
fp_data = DATASETS[dataset_code]['data-fp']
data = self.sdmx.get(
resource_type='data', agency='FR1', fromfile=fp_data)
series = list(data.msg.data.series)
series = series[0]
self.assertEqual(list(series.key._asdict().keys()),
['SECT-INST', 'OPERATION', 'PRODUIT', 'PRIX'])
self.assertEqual(list(series.attrib._asdict().keys()),
['FREQ', 'IDBANK', 'TITLE', 'LAST_UPDATE', 'UNIT_MEASURE', 'UNIT_MULT', 'REF_AREA', 'DECIMALS', 'BASE_PER', 'TIME_PER_COLLECT'])
def test_freq_in_series_attribute(self):
# Test that we don't have regression on Issues #39 and #41
# INSEE time series provide the FREQ value as attribute on the series instead of a dimension. This caused
# a runtime error when writing as pandas dataframe.
data_response = self.sdmx.data(
fromfile=SERIES['UNEMPLOYMENT_CAT_A_B_C']['data-fp'])
data_response.write()
class InseeTestCase(unittest.TestCase):
# nosetests -s -v pandasdmx.tests.test_insee:InseeTestCase
def setUp(self):
unittest.TestCase.setUp(self)
self.sdmx = Request('INSEE')
def test_load_dataset(self):
dataset_code = 'IPI-2010-A21'
'''load all dataflows'''
dataflows_response = self.sdmx.get(
resource_type='dataflow', agency='FR1', fromfile=DATAFLOW_FP)
dataflows = dataflows_response.msg.dataflow
self.assertEqual(len(dataflows.keys()), 663)
self.assertTrue(dataset_code in dataflows)
'''load datastructure for current dataset_code'''
fp_datastructure = DATASETS[dataset_code]['datastructure-fp']
datastructure_response = self.sdmx.get(
resource_type='datastructure', agency='FR1', fromfile=fp_datastructure)
self.assertTrue(
dataset_code in datastructure_response.msg.datastructure)
dsd = datastructure_response.msg.datastructure[dataset_code]
'''Verify dimensions list'''
dimensions = OrderedDict([dim.id, dim] for dim in dsd.dimensions.aslist(
) if dim.id not in ['TIME', 'TIME_PERIOD'])
dim_keys = list(dimensions.keys())
self.assertEqual(dim_keys, ['FREQ', 'PRODUIT', 'NATURE'])
'''load datas for the current dataset'''
fp_data = DATASETS[dataset_code]['data-fp']
data = self.sdmx.get(
resource_type='data', agency='FR1', fromfile=fp_data)
'''Verify series count and values'''
series = list(data.msg.data.series)
series_count = len(series)
self.assertEqual(series_count, DATASETS[dataset_code]['series_count'])
first_series = series[0]
observations = list(first_series.obs())
first_obs = observations[0]
last_obs = observations[-1]
self.assertEqual(first_obs.dim, '2015-10')
self.assertEqual(first_obs.value, '105.61')
self.assertEqual(last_obs.dim, '1990-01')
self.assertEqual(last_obs.value, '139.22')
def test_fixe_key_names(self):
"""Verify key or attribute contains '-' in name
"""
dataset_code = 'CNA-2010-CONSO-SI-A17'
fp_datastructure = DATASETS[dataset_code]['datastructure-fp']
datastructure_response = self.sdmx.get(
resource_type='datastructure', agency='FR1', fromfile=fp_datastructure)
self.assertTrue(
dataset_code in datastructure_response.msg.datastructure)
dsd = datastructure_response.msg.datastructure[dataset_code]
dimensions = OrderedDict([dim.id, dim] for dim in dsd.dimensions.aslist(
) if dim.id not in ['TIME', 'TIME_PERIOD'])
dim_keys = list(dimensions.keys())
self.assertEqual(
dim_keys, ['SECT-INST', 'OPERATION', 'PRODUIT', 'PRIX'])
fp_data = DATASETS[dataset_code]['data-fp']
data = self.sdmx.get(
resource_type='data', agency='FR1', fromfile=fp_data)
series = list(data.msg.data.series)
series = series[0]
self.assertEqual(list(series.key._asdict().keys()),
['SECT-INST', 'OPERATION', 'PRODUIT', 'PRIX'])
self.assertEqual(list(series.attrib._asdict().keys()),
['FREQ', 'IDBANK', 'TITLE', 'LAST_UPDATE', 'UNIT_MEASURE', 'UNIT_MULT', 'REF_AREA', 'DECIMALS', 'BASE_PER', 'TIME_PER_COLLECT'])
def setUp(self):
unittest.TestCase.setUp(self)
self.sdmx = Request('INSEE')
def req(self):
return Request("INSEE")
def setUp(self):
self.estat = Request('ESTAT')
filepath = os.path.join(test_path, 'data/common/common.xml')
self.resp = self.estat.get(fromfile=filepath)
def setUp(self):
self.estat = Request('ESTAT')
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_ng/generic/ecb_exr_ng_ts_gf.xml')
self.resp = self.estat.data(fromfile=filepath)
"""
Created on Fri Oct 19 10:02:37 2018
@author: dpsugasa
"""
from pandasdmx import Request
import plotly
import plotly.plotly as py #for plotting
import plotly.graph_objs as go
import plotly.dashboard_objs as dashboard
import plotly.tools as tls
import plotly.figure_factory as ff
import credentials #plotly API details
estat = Request('ESTAT')
flow_response = estat.dataflow('une_rt_a')
structure_response = flow_response.dataflow.une_rt_a.structure(
request=True, target_only=False)
structure_response.write().codelist.loc['GEO'].head()
resp = estat.data('une_rt_a',
key={'GEO': 'EL+ES+IT'},
params={'startPeriod': '1950'})
data = resp.write(s for s in resp.data.series if s.key.AGE == 'TOTAL')
z = data.loc[:, ('PC_ACT', 'TOTAL', 'T')]
z['Dates'] = z.index.to_series().astype(str)
z['IT_scr'] = (z['IT'] - z['IT'].mean()) / z['IT'].std()
print(z)
class EurostatFetcher(SDMXFetcher):
_sdmx_loader = Request('ESTAT')
_nsmap = dict(nt='urn:eu.europa.ec.eurostat.navtree')
_GEO = ['AT', 'BE', 'BG', 'CY', 'CZ', 'DE', 'DK', 'EE', 'EL', 'ES',
'EU28', 'EU27', 'EA19', 'FI', 'FR', 'HR', 'HU', 'IE', 'IT', 'LT', 'LU', 'LV', 'ME',
'MK', 'MT', 'NL', 'PL', 'PT', 'RO', 'RS', 'SE', 'SI', 'SK', 'TR', 'UK']
_S_ADJ = ['NSA','SA']
_FREQ = ['A','H','Q','M','S','W']
def __init__(self, name=NAME, parent_location=GIT_PARENT_LOCATION):
self._name = name
self._base_url = URL
self._parent_location = parent_location
self._repo_location = parent_location + "/" + name
self._temp_location = TEMP_PARENT_LOCATION + "/" + name
self._xml_toc_file = XML_TOC_FILE
self._xml_toc_file_location = self._repo_location + "/" + self._xml_toc_file
Path(self._temp_location).mkdir(exist_ok=True)
Path(self._repo_location).mkdir(exist_ok=True)
try:
self.load_existing_repo()
except ValueError:
self.initialize_repo()
def _download_file(self, params, stream):
rsp = requests.get(self._base_url,params,stream=stream)
if rsp.ok:
file_path = self._repo_location + '/' + params['file'].split('/')[-1]
with open(file_path, 'wb') as f:
for chunk in rsp.iter_content(chunk_size=1024):
if chunk:
f.write(chunk)
f.close()
self.repo.stage([file_path])
return rsp
def _dictionary_file(self, params):
rsp = self._download_file(params, False)
if rsp.ok:
tsv_file_io = StringIO(rsp.text)
reader = csv.reader(tsv_file_io,delimiter='\t')
return dict([(r[0],r[1]) for r in reader])
def _dimensions_list(self):
log.info("Downloading and updating the dimensions list...")
params = {"sort":"1","file":"dic/en/dimlst.dic"}
self._dimlst = self._dictionary_file(params)
def _load_existing_toc(self):
if Path(self._xml_toc_file_location).is_file():
log.info("Loading up existing Table of Contents XML file...")
parser = etree.XMLParser(remove_blank_text=True)
self.xml_toc_old = etree.parse(self._xml_toc_file_location, parser=parser)
else:
log.info("No existing Table of Contents XML file. Setting to None.")
self.xml_toc_old = None
def _update_toc(self):
self._load_existing_toc()
log.info("Starting file Table of Contents XML file download...")
response = requests.get(XML_TOC_URL)
log.info(">> download finished.")
parser = etree.XMLParser(remove_blank_text=True)
xml_toc = etree.fromstring(response.content, parser=parser)
self.xml_toc = etree.ElementTree(xml_toc)
with open(self._xml_toc_file_location, 'wb') as xml_file:
self.xml_toc.write(xml_file, encoding='utf-8', pretty_print=True, xml_declaration=True)
xml_file.close()
self.repo.stage([self._xml_toc_file])
def _datasets_to_update(self, forceUpdateAll=False):
if forceUpdateAll:
for ds in self.xml_toc.iterfind('.//nt:leaf[@type="dataset"]', namespaces=self._nsmap):
yield ds
else:
if self.xml_toc_old is not None:
last_update_by_ds = {}
for ds in self.xml_toc_old.iterfind('.//nt:leaf[@type="dataset"]', namespaces=self._nsmap):
ds_code = ds.findtext("nt:code", namespaces=self._nsmap)
ds_last_update = ds.findtext("nt:lastUpdate", namespaces=self._nsmap)
last_update_by_ds[ds_code] = ds_last_update
for ds in self.xml_toc.iterfind('.//nt:leaf[@type="dataset"]', namespaces=self._nsmap):
ds_code = ds.findtext("nt:code", namespaces=self._nsmap)
if self.xml_toc_old is None:
yield ds
else:
ds_earlier_update = last_update_by_ds.get(ds_code)
if ds_earlier_update is None:
yield ds
else:
ds_last_update = ds.findtext("nt:lastUpdate", namespaces=self._nsmap)
if ds_last_update != ds_earlier_update:
yield ds
def update_content(self, forceUpdateAll=False):
ds_iter = self._datasets_to_update(forceUpdateAll)
for ds in ds_iter:
ds_code = ds.findtext("nt:code", namespaces=self._nsmap)
#ds_last_update = ds.findtext("nt:lastUpdate", namespaces=self._nsmap)
#ds_metadata = ds.findtext('nt:metadata[@format="sdmx"]', namespaces=self._nsmap)
#ds_url_tsv = ds.findtext('nt:downloadLink[@format="tsv"]', namespaces=self._nsmap)
#ds_url_sdmx = ds.findtext('nt:downloadLink[@format="sdmx"]', namespaces=self._nsmap)
log.info('Dowloading dataset {}'.format(ds_code))
#SHORT APPROACH TO DOWNLOAD DATA AS TSV ONLY HAS SERIES
#log.info('...tsv file...')
#response = requests.get(ds_url_tsv)
#LONG APPROACH TO DOWNLOAD DATA AS SDMX-->VERY INEFFICIENT IT SEEMS
#BETTER TO DOWNLOAD IN ZIP, THEN UNZIP AND THEN READ FROM FILES?
log.info('...sdmx information...'.format(ds_code))
dsd_code = 'DSD_' + ds_code
cs_code = 'CS_' + dsd_code
dsd_response = self._sdmx_loader.datastructure(resource_id=dsd_code)
dsd = dsd_response.msg.datastructure[dsd_code]
dsd_df = dsd_response.write()
cs = dsd_response.msg.conceptscheme[cs_code]
cs_id_list = [c.id for c in cs.aslist()]
cl = dsd_response.msg.codelist
data_response = self._sdmx_loader.data(resource_id=ds_code, dsd=dsd)
data = data_response.data
series_l = list(data.series)
series_df = data_response.write(series_l)
#this is because the timeseries are in reverse order otherwise
series_df.sort_index(inplace=True)
idx_multi = series_df.columns
#drop levels that are unique in the multi-index
lvl_to_drop = [i for i in range(len(idx_multi.levels)) if len(idx_multi.levels[i])==1]
idx_multi = idx_multi.droplevel(lvl_to_drop)
if 'INDIC' in idx_multi.names:
idx_multi.swaplevel('INDIC',0) #place INDIC first
if 'INDICATOR' in idx_multi.names:
idx_multi.swaplevel('INDICATOR',0) #place INDICATOR first
series_df.columns = idx_multi
idx_multi_joined = ['_'.join(col).strip() for col in idx_multi.values]
series_df.columns = idx_multi_joined
series_df.to_csv(self._repo_location + ds_code +'.csv', na_rep="na")
def setUp(self):
unittest.TestCase.setUp(self)
self.sdmx = Request('INSEE')
def setUp(self):
self.estat = Request('ESTAT')
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_ng/generic/ecb_exr_ng_flat.xml')
self.resp = self.estat.get(fromfile=filepath)
def test_doc_usage_structure():
"""Code examples in walkthrough.rst."""
ecb = Request("ECB")
ecb_via_proxy = Request("ECB", proxies={"http": "http://1.2.3.4:5678"})
assert all(
getattr(ecb_via_proxy.session, k) == v
for k, v in (
("proxies", {"http": "http://1.2.3.4:5678"}),
("stream", False),
("timeout", 30.1),
)
)
msg1 = ecb.categoryscheme()
assert msg1.response.url == (
"https://sdw-wsrest.ecb.europa.eu/service/categoryscheme/ECB/latest"
"?references=parentsandsiblings"
)
# Check specific headers
headers = msg1.response.headers
assert headers["Content-Type"] == (
"application/vnd.sdmx.structure+xml; " "version=2.1"
)
assert all(k in headers for k in ["Connection", "Date", "Server"])
# Removed: in pandaSDMX 0.x this was a convenience method that (for this
# structure message) returned two DataStructureDefinitions. Contra the
# spec, that assumes:
# - There is 1 Categorization using the CategoryScheme; there could be
# many.
# - The Categorization maps DataStructureDefinitions to Categories, when
# there could be many.
# list(cat_response.category_scheme['MOBILE_NAVI']['07'])
dfs = pandasdmx.to_pandas(msg1.dataflow).head()
assert len(dfs) == 2
flows = ecb.dataflow() # noqa: F841
dsd_id = flows.dataflow.EXR.structure.id
assert dsd_id == "ECB_EXR1"
refs = dict(references="all")
msg2 = ecb.datastructure(resource_id=dsd_id, params=refs)
dsd = msg2.structure[dsd_id]
assert pandasdmx.to_pandas(dsd.dimensions) == [
"FREQ",
"CURRENCY",
"CURRENCY_DENOM",
"EXR_TYPE",
"EXR_SUFFIX",
"TIME_PERIOD",
]
cl = pandasdmx.to_pandas(msg2.codelist["CL_CURRENCY"]).sort_index()
expected = pd.Series(
{
"ADF": "Andorran Franc (1-1 peg to the French franc)",
"ADP": "Andorran Peseta (1-1 peg to the Spanish peseta)",
"AED": "United Arab Emirates dirham",
"AFA": "Afghanistan afghani (old)",
"AFN": "Afghanistan, Afghanis",
},
name="Currency code list",
).rename_axis("CL_CURRENCY")
assert_pd_equal(cl.head(), expected)
class TestStructSpecSeriesDataSet(unittest.TestCase):
def setUp(self):
self.estat = Request('ESTAT')
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_ng/structured/ecb_exr_ng_ts_gf.xml')
dsd_resp = self.estat.datastructure(
fromfile=os.path.join(pkg_path, 'data/exr/ecb_exr_ng/ecb_exr_ng_full.xml'))
dsd = dsd_resp.datastructure.DataStructure
self.resp = self.estat.data(fromfile=filepath, dsd=dsd)
def test_header_attributes(self):
self.assertEqual(self.resp.header.structured_by, 'STR1')
self.assertEqual(self.resp.header.dim_at_obs, 'TIME_PERIOD')
def test_dataset_cls(self):
self.assertIsInstance(self.resp.msg.data, model.DataSet)
def test_obs(self):
data = self.resp.data
# empty obs iterator
self.assertEqual(len(list(data.obs())), 0)
series_list = list(data.series)
self.assertEqual(len(series_list), 4)
s3 = series_list[3]
self.assertIsInstance(s3, model.Series)
self.assertIsInstance(s3.key, tuple)
self.assertEqual(len(s3.key), 4)
self.assertEqual(s3.key.CURRENCY, 'USD')
self.assertEqual(s3.attrib.DECIMALS, '4')
obs_list = list(s3.obs(reverse_obs=True))
self.assertEqual(len(obs_list), 3)
o0 = obs_list[2]
self.assertEqual(len(o0), 3)
self.assertEqual(o0.dim, '2010-08')
self.assertEqual(o0.value, '1.2894')
self.assertIsInstance(o0.attrib, tuple)
self.assertEqual(o0.attrib.OBS_STATUS, 'A')
def test_pandas(self):
resp = self.resp
data = resp.data
pd_series = [s for s in resp.write(
data, attributes='', reverse_obs=True, asframe=False)]
self.assertEqual(len(pd_series), 4)
s3 = pd_series[3]
self.assertIsInstance(s3, pandas.core.series.Series)
self.assertEqual(s3[2], 1.2894)
self.assertIsInstance(s3.name, tuple)
self.assertEqual(len(s3.name), 4)
# now with attributes
pd_series = [s for s in resp.write(
data, attributes='osgd', reverse_obs=True, asframe=False)]
self.assertEqual(len(pd_series), 4)
self.assertIsInstance(pd_series[0], tuple) # contains 2 series
self.assertEqual(len(pd_series[0]), 2)
s3, a3 = pd_series[3]
self.assertIsInstance(s3, pandas.core.series.Series)
self.assertIsInstance(a3, pandas.core.series.Series)
self.assertEqual(s3[2], 1.2894)
self.assertIsInstance(s3.name, tuple)
self.assertEqual(len(s3.name), 4)
self.assertEqual(len(a3), 3)
# access an attribute of the first value
self.assertEqual(a3[0].OBS_STATUS, 'A')
def test_write2pandas(self):
df = self.resp.write(attributes='',
reverse_obs=False)
self.assertIsInstance(df, pandas.DataFrame)
assert df.shape == (3, 4)
# with metadata
df, mdf = self.resp.write(attributes='osgd',
reverse_obs=False)
assert mdf.shape == (3, 4)
assert mdf.iloc[1, 1].OBS_STATUS == 'A'
def istat():
return Request("ISTAT")
#!/usr/bin/env python
# coding: utf-8
import math, re, sys, calendar, os, copy, time
import pandas as pd
import numpy as np
from datetime import datetime, date
# In[80]:
import pandasdmx
from pandasdmx import Request
tStart = time.time()
oecd = Request('OECD')
print('Time: ', int(time.time() - tStart),'s'+'\n')
data_response = oecd.data(resource_id='MEI_BTS_COS', key='all?startTime=1995')
print('Time: ', int(time.time() - tStart),'s'+'\n')
df = data_response.to_pandas()
print('Time: ', int(time.time() - tStart),'s'+'\n')
#df.to_csv('c:\\Temp\\test_lei.txt', sep='\t')
path = 'C:/Users/lawre/Desktop'
# In[82]:
#df = data_response.to_pandas()
#type(data_response.data)
df.to_csv(path+'test_lei.csv')
def setUp(self):
self.estat = Request()
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_rg/generic/ecb_exr_rg_ts.xml')
self.resp = self.estat.get(fromfile=filepath)
def test_doc_usage_structure():
"""Code examples in walkthrough.rst."""
ecb = Request('ECB')
ecb_via_proxy = Request('ECB', proxies={'http': 'http://1.2.3.4:5678'})
assert all(getattr(ecb_via_proxy.session, k) == v for k, v in (
('proxies', {'http': 'http://1.2.3.4:5678'}),
('stream', True),
('timeout', 30.1),
))
msg1 = ecb.categoryscheme()
assert msg1.response.url == (
'http://sdw-wsrest.ecb.int/service/categoryscheme/ECB/latest'
'?references=parentsandsiblings')
# Check specific headers
headers = msg1.response.headers
assert headers['Content-Type'] == ('application/vnd.sdmx.structure+xml; '
'version=2.1')
assert all(k in headers for k in ['Connection', 'Date', 'Server'])
# Removed: in pandaSDMX 0.x this was a convenience method that (for this
# structure message) returned two DataStructureDefinitions. Contra the
# spec, that assumes:
# - There is 1 Categorization using the CategoryScheme; there could be many.
# - The Categorization maps DataStructureDefintions to Categories, when
# there could be many.
# list(cat_response.category_scheme['MOBILE_NAVI']['07'])
dfs = sdmx.to_pandas(msg1.dataflow).head()
expected = pd.Series({
'AME': 'AMECO',
'BKN': 'Banknotes statistics',
'BLS': 'Bank Lending Survey Statistics',
'BOP': ('Euro Area Balance of Payments and International Investment '
'Position Statistics'),
'BSI': 'Balance Sheet Items',
})
assert_pd_equal(dfs, expected)
flows = ecb.dataflow()
dsd_id = msg1.dataflow.EXR.structure.id
assert dsd_id == 'ECB_EXR1'
refs = dict(references='all')
msg2 = ecb.datastructure(resource_id=dsd_id, params=refs)
dsd = msg2.structure[dsd_id]
assert sdmx.to_pandas(dsd.dimensions) == ['FREQ', 'CURRENCY',
'CURRENCY_DENOM', 'EXR_TYPE', 'EXR_SUFFIX', 'TIME_PERIOD']
cl = sdmx.to_pandas(msg2.codelist['CL_CURRENCY']).sort_index()
expected = pd.Series({
'ADF': 'Andorran Franc (1-1 peg to the French franc)',
'ADP': 'Andorran Peseta (1-1 peg to the Spanish peseta)',
'AED': 'United Arab Emirates dirham',
'AFA': 'Afghanistan afghani (old)',
'AFN': 'Afghanistan, Afghanis',
}, name='Currency code list') \
.rename_axis('CL_CURRENCY')
assert_pd_equal(cl.head(), expected)
def setUp(self):
self.estat = Request()
filepath = os.path.join(
pkg_path, 'data/exr/ecb_exr_sg/generic/ecb_exr_sg_ts.xml')
self.resp = self.estat.get(fromfile=filepath)
from pandasdmx import Request
# Connecting to the Eurostat web service
estat = Request('ESTAT')
# Downloading the dataflow definitions
flows = estat.dataflow()
# Getting information about the dataflow
flows.url
flows.http_headers
# Exporting the dataflow definitions to a pandas DataFrame
dflows = flows.write().dataflow
# Listing tables from the high tech database and the description of a single table
ht_tabs = dflows[dflows.index.str.startswith('htec') == True]
kia_emp = dflows.loc['htec_kia_emp2'][0]
# Dataflow definition
df_def = flows.dataflow.htec_kia_emp2
# Database's datastructure id
dsd_id = df_def.structure.id
# Creating a support dict
refs = dict(references = 'all')
# Calling the table
dsd_response = estat.get(url = 'http://ec.europa.eu/eurostat/SDMX/diss-web/rest/datastructure/ESTAT/' + dsd_id)
# Getting informatou about the datastructure
