def process(self):
self.logger.setLevel(self._log_levels[self.args.verbose])
ad = ArgoData(verbosity=self.args.verbose, cache_file=self.args.cache_file)
if self.args.wmo:
wmo_list = self.args.wmo
else:
wmo_list = ad.get_cache_file_oxy_count_df()["wmo"].tolist()
self.logger.info("Reading float profile data from %s", self.args.cache_file)
for i, wmo in enumerate(wmo_list):
self.logger.info("WMO_%s: Float %s of %s", wmo, i + 1, len(wmo_list))
try:
with pd.HDFStore(self.args.results_file) as s:
self.logger.debug("Reading %s from %s", ("/WOA_WMO_{}").format(wmo), self.args.cache_file)
wmo_gdf = s.get(("/WOA_WMO_{}").format(wmo))
self.logger.debug("Done.")
except KeyError:
df = ad.get_float_dataframe(
[wmo], max_profiles=self.args.profiles, max_pressure=self.args.pressure, update_cache=False
)
wmo_gdf = self.woa_lookup(df)
# Save intermediate results to HDF file so that the script can
# pick up where it left off following network or other problems
with pd.HDFStore(self.args.results_file) as s:
s.put(("/WOA_WMO_{}").format(wmo), wmo_gdf)
if not wmo_gdf.dropna().empty:
self.logger.debug("wmo_gdf head: %s", wmo_gdf.head())
self.logger.info("Gain for %s = %s", wmo, wmo_gdf.groupby("wmo").gain.mean().values[0])
def process(self):
if self.args.cache_file:
cache_file = self.args.cache_file
else:
cache_file = self.short_cache_file()
if self.args.cache_dir:
cache_dir = self.args.cache_dir
else:
cache_dir = expanduser('~')
cache_file = abspath(join(cache_dir, cache_file))
print(('Loading cache file {}').format(cache_file))
ad = ArgoData(verbosity=self.args.verbose, cache_file=cache_file,
bio_list=self.args.bio_list, variables=self.args.variables)
if self.args.age:
wmo_list = ad.get_oxy_floats_from_status(age_gte=self.args.age)
elif self.args.wmo:
wmo_list = self.args.wmo
ad.get_float_dataframe(wmo_list, max_profiles=self.args.profiles,
max_pressure=self.args.pressure,
append_df=False)
# After loading add lookup information to the cache file
df = ad.get_cache_file_oxy_count_df(max_profiles=self.args.profiles, flush=True)
print(('{} floats appear to have valid oxygen data').format(len(df)))
print(('Finished loading cache file {}').format(cache_file))
def process(self):
self.logger.setLevel(self._log_levels[self.args.verbose])
ad = ArgoData(verbosity=self.args.verbose,
cache_file=self.args.cache_file)
if self.args.wmo:
wmo_list = self.args.wmo
else:
wmo_list = ad.get_cache_file_oxy_count_df()['wmo'].tolist()
self.logger.info('Reading float profile data from %s',
self.args.cache_file)
for i, wmo in enumerate(wmo_list):
self.logger.info('WMO_%s: Float %s of %s', wmo, i + 1,
len(wmo_list))
try:
with pd.HDFStore(self.args.results_file) as s:
self.logger.debug('Reading %s from %s',
('/WOA_WMO_{}').format(wmo),
self.args.cache_file)
wmo_gdf = s.get(('/WOA_WMO_{}').format(wmo))
self.logger.debug('Done.')
except KeyError:
df = ad.get_float_dataframe([wmo],
max_profiles=self.args.profiles,
max_pressure=self.args.pressure,
update_cache=False)
wmo_gdf = self.woa_lookup(df)
if not wmo_gdf.dropna().empty:
# Save intermediate results to HDF file so that the script can
# pick up where it left off following network or other problems
with pd.HDFStore(self.args.results_file) as s:
s.put(('/WOA_WMO_{}').format(wmo), wmo_gdf)
else:
self.logger.warn('Empty DataFrame for wmo %s', wmo)
if not wmo_gdf.dropna().empty:
self.logger.debug('wmo_gdf head: %s', wmo_gdf.head())
self.logger.info('Gain for %s = %s', wmo,
wmo_gdf.groupby('wmo').gain.mean().values[0])
def process(self):
if self.args.cache_file:
cache_file = self.args.cache_file
else:
cache_file = self.short_cache_file()
if self.args.cache_dir:
cache_dir = self.args.cache_dir
else:
cache_dir = expanduser('~')
cache_file = abspath(join(cache_dir, cache_file))
print(('Loading cache file {}').format(cache_file))
ad = ArgoData(verbosity=self.args.verbose,
cache_file=cache_file,
bio_list=self.args.bio_list,
variables=self.args.variables)
if self.args.age:
wmo_list = ad.get_oxy_floats_from_status(age_gte=self.args.age)
elif self.args.wmo:
wmo_list = self.args.wmo
ad.get_float_dataframe(wmo_list,
max_profiles=self.args.profiles,
max_pressure=self.args.pressure,
append_df=False)
# After loading add lookup information to the cache file
df = ad.get_cache_file_oxy_count_df(max_profiles=self.args.profiles,
flush=True)
print(('{} floats appear to have valid oxygen data').format(len(df)))
print(('Finished loading cache file {}').format(cache_file))
def test_fixed_cache_file(self):
age = 3000 # Returns 1 float on 2 November 2015
parent_dir = os.path.join(os.path.dirname(__file__), "../")
# Simulated what's done by load_biofloat_cache.py
from scripts.load_biofloat_cache import ArgoDataLoader
from argparse import Namespace
adl = ArgoDataLoader()
adl.args = Namespace(age=3000, profiles=1)
cache_file = os.path.abspath(
os.path.join(parent_dir, 'biofloat', adl.short_cache_file()))
ad = ArgoData(verbosity=1, cache_file=cache_file)
wmo_list = ad.get_oxy_floats_from_status(age_gte=age)
# Force limiting to what's in cache_file name: 1
ad.get_float_dataframe(wmo_list, max_profiles=2)
# Force using maximum value
ad.get_float_dataframe(wmo_list)
def test_fixed_cache_file(self):
age = 3000 # Returns 1 float on 2 November 2015
parent_dir = os.path.join(os.path.dirname(__file__), "../")
# Simulated what's done by load_biofloat_cache.py
from scripts.load_biofloat_cache import ArgoDataLoader
from argparse import Namespace
adl = ArgoDataLoader()
adl.args = Namespace(age=3000, profiles=1)
cache_file = os.path.abspath(
os.path.join(parent_dir, 'biofloat', adl.short_cache_file()))
ad = ArgoData(verbosity=1, cache_file=cache_file)
wmo_list = ad.get_oxy_floats_from_status(age_gte=age)
# Force limiting to what's in cache_file name: 1
ad.get_float_dataframe(wmo_list, max_profiles=2)
# Force using maximum value
ad.get_float_dataframe(wmo_list)
def test_cache_file(self):
ad = ArgoData(cache_file='/tmp/biofloat_cache_file.hdf')
ad.set_verbosity(1)
def setUp(self):
self.ad = ArgoData(verbosity=1)
self.good_oga_floats = ['1900650']
self.bad_oga_floats = ['6901464']
self._build_default_cache()
class DataTest(unittest.TestCase):
def setUp(self):
self.ad = ArgoData(verbosity=1)
self.good_oga_floats = ['1900650']
self.bad_oga_floats = ['6901464']
self._build_default_cache()
def test_get_biofloats(self):
self.oga_floats = self.ad.get_oxy_floats_from_status()
self.assertNotEqual(len(self.oga_floats), 0)
def _get_dac_urls(self):
# Testing with a float that has data
for dac_url in self.ad.get_dac_urls(self.good_oga_floats).values():
self.dac_url = dac_url
self.assertTrue(self.dac_url.startswith('http'))
break
def _get_profile_opendap_urls(self):
for profile_url in self.ad.get_profile_opendap_urls(self.dac_url):
self.profile_url = profile_url
break
def _profile_to_dataframe(self):
key, code = self.ad._float_profile_key(self.profile_url)
d = self.ad._profile_to_dataframe(self.good_oga_floats[0],
self.profile_url, key, 11000)
self.assertNotEqual(len(d), 0)
def _build_default_cache(self):
# Methods need to be called in order
self._get_dac_urls()
self._get_profile_opendap_urls()
self._profile_to_dataframe()
def test_get_float_dataframe(self):
df = self.ad.get_float_dataframe(self.good_oga_floats, max_profiles=2)
self.assertNotEqual(len(df), 0)
df = self.ad.get_float_dataframe(self.bad_oga_floats, max_profiles=2)
self.assertEqual(len(df.dropna()), 0)
def test_cache_file(self):
ad = ArgoData(cache_file='/tmp/biofloat_cache_file.hdf')
ad.set_verbosity(1)
def test_fixed_cache_file(self):
age = 3000 # Returns 1 float on 2 November 2015
parent_dir = os.path.join(os.path.dirname(__file__), "../")
# Simulated what's done by load_biofloat_cache.py
from scripts.load_biofloat_cache import ArgoDataLoader
from argparse import Namespace
adl = ArgoDataLoader()
adl.args = Namespace(age=3000, profiles=1)
cache_file = os.path.abspath(
os.path.join(parent_dir, 'biofloat', adl.short_cache_file()))
ad = ArgoData(verbosity=1, cache_file=cache_file)
wmo_list = ad.get_oxy_floats_from_status(age_gte=age)
# Force limiting to what's in cache_file name: 1
ad.get_float_dataframe(wmo_list, max_profiles=2)
# Force using maximum value
ad.get_float_dataframe(wmo_list)
def test_util_o2sat(self):
# See http://www.engineeringtoolbox.com/oxygen-solubility-water-d_841.html
self.assertAlmostEqual(utils.o2sat(35, 5), 308, places=0)
self.assertAlmostEqual(utils.o2sat(35, 20), 225, places=0)
self.assertAlmostEqual(utils.o2sat(35, 30), 190, places=0)
def test_util_convert_to_mll(self):
# See http://www.engineeringtoolbox.com/oxygen-solubility-water-d_841.html
# and http://www.oceanographers.net/forums/showthread.php?1486-ask-how-to-conversion-ml-L-to-%B5mol-kg
self.assertAlmostEqual(utils.convert_to_mll(308, 35, 5, 0),
7.1,
places=1)
self.assertAlmostEqual(utils.convert_to_mll(225.6, 36.5, 1, 0),
5.2,
places=1)
def test_get_bio_profile_index(self):
df = self.ad.get_bio_profile_index()
self.assertNotEqual(len(df), 0)
def test_get_cache_file_all_wmo_list(self):
df = self.ad.get_float_dataframe(self.good_oga_floats, max_profiles=2)
self.assertNotEqual(len(df), 0)
df = self.ad.get_cache_file_all_wmo_list(flush=True)
self.assertNotEqual(len(df), 0)
df = self.ad.get_cache_file_all_wmo_list()
self.assertNotEqual(len(df), 0)
def test_get_cache_file_oxy_wmo_list(self):
df = self.ad.get_float_dataframe(self.good_oga_floats, max_profiles=2)
self.assertNotEqual(len(df), 0)
df = self.ad.get_cache_file_oxy_count_df(max_profiles=2, flush=True)
self.assertNotEqual(len(df), 0)
df = self.ad.get_cache_file_oxy_count_df(max_profiles=2)
self.assertNotEqual(len(df), 0)
def test_get_update_datetime(self):
# Any random delayed mode file will do...
dt = self.ad._get_update_datetime(
'http://tds0.ifremer.fr/thredds/dodsC/CORIOLIS-ARGO-GDAC-OBS/aoml/5901336/profiles/D5901336_082.nc'
)
def test_to_odv(self):
df = self.ad.get_float_dataframe(self.good_oga_floats, max_profiles=2)
converters.to_odv(df, 'biofloat_data.txt')
def test_update_cache_false(self):
df = self.ad.get_float_dataframe(self.good_oga_floats,
max_profiles=2,
update_cache=False)
self.assertNotEqual(len(df), 0)
def test_remove_df(self):
self.ad._remove_df(self.ad._GLOBAL_META)
def test_cache_file(self):
ad = ArgoData(cache_file='/tmp/biofloat_cache_file.hdf')
ad.set_verbosity(1)
def setUp(self):
self.ad = ArgoData(verbosity=1)
self.good_oga_floats = ['1900650']
self.bad_oga_floats = ['6901464']
self._build_default_cache()
class DataTest(unittest.TestCase):
def setUp(self):
self.ad = ArgoData(verbosity=1)
self.good_oga_floats = ['1900650']
self.bad_oga_floats = ['6901464']
self._build_default_cache()
def test_get_biofloats(self):
self.oga_floats = self.ad.get_oxy_floats_from_status()
self.assertNotEqual(len(self.oga_floats), 0)
def _get_dac_urls(self):
# Testing with a float that has data
for dac_url in self.ad.get_dac_urls(self.good_oga_floats).values():
self.dac_url = dac_url
self.assertTrue(self.dac_url.startswith('http'))
break
def _get_profile_opendap_urls(self):
for profile_url in self.ad.get_profile_opendap_urls(self.dac_url):
self.profile_url = profile_url
break
def _profile_to_dataframe(self):
key, code = self.ad._float_profile_key(self.profile_url)
d = self.ad._profile_to_dataframe(self.good_oga_floats[0],
self.profile_url, key, 11000)
self.assertNotEqual(len(d), 0)
def _build_default_cache(self):
# Methods need to be called in order
self._get_dac_urls()
self._get_profile_opendap_urls()
self._profile_to_dataframe()
def test_get_float_dataframe(self):
df = self.ad.get_float_dataframe(self.good_oga_floats, max_profiles=2)
self.assertNotEqual(len(df), 0)
df = self.ad.get_float_dataframe(self.bad_oga_floats, max_profiles=2)
self.assertEqual(len(df.dropna()), 0)
def test_cache_file(self):
ad = ArgoData(cache_file='/tmp/biofloat_cache_file.hdf')
ad.set_verbosity(1)
def test_fixed_cache_file(self):
age = 3000 # Returns 1 float on 2 November 2015
parent_dir = os.path.join(os.path.dirname(__file__), "../")
# Simulated what's done by load_biofloat_cache.py
from scripts.load_biofloat_cache import ArgoDataLoader
from argparse import Namespace
adl = ArgoDataLoader()
adl.args = Namespace(age=3000, profiles=1)
cache_file = os.path.abspath(
os.path.join(parent_dir, 'biofloat', adl.short_cache_file()))
ad = ArgoData(verbosity=1, cache_file=cache_file)
wmo_list = ad.get_oxy_floats_from_status(age_gte=age)
# Force limiting to what's in cache_file name: 1
ad.get_float_dataframe(wmo_list, max_profiles=2)
# Force using maximum value
ad.get_float_dataframe(wmo_list)
def test_util_o2sat(self):
# See http://www.engineeringtoolbox.com/oxygen-solubility-water-d_841.html
self.assertAlmostEqual(utils.o2sat(35, 5), 308, places=0)
self.assertAlmostEqual(utils.o2sat(35, 20), 225, places=0)
self.assertAlmostEqual(utils.o2sat(35, 30), 190, places=0)
def test_util_convert_to_mll(self):
# See http://www.engineeringtoolbox.com/oxygen-solubility-water-d_841.html
# and http://www.oceanographers.net/forums/showthread.php?1486-ask-how-to-conversion-ml-L-to-%B5mol-kg
self.assertAlmostEqual(utils.convert_to_mll(308, 35, 5, 0), 7.1, places=1)
self.assertAlmostEqual(utils.convert_to_mll(225.6, 36.5, 1, 0), 5.2, places=1)
def test_get_bio_profile_index(self):
df = self.ad.get_bio_profile_index()
self.assertNotEqual(len(df), 0)
def test_get_cache_file_all_wmo_list(self):
df = self.ad.get_float_dataframe(self.good_oga_floats, max_profiles=2)
self.assertNotEqual(len(df), 0)
df = self.ad.get_cache_file_all_wmo_list(flush=True)
self.assertNotEqual(len(df), 0)
df = self.ad.get_cache_file_all_wmo_list()
self.assertNotEqual(len(df), 0)
def test_get_cache_file_oxy_wmo_list(self):
df = self.ad.get_float_dataframe(self.good_oga_floats, max_profiles=2)
self.assertNotEqual(len(df), 0)
df = self.ad.get_cache_file_oxy_count_df(max_profiles=2, flush=True)
self.assertNotEqual(len(df), 0)
df = self.ad.get_cache_file_oxy_count_df(max_profiles=2)
self.assertNotEqual(len(df), 0)
def test_get_update_datetime(self):
# Any random delayed mode file will do...
dt = self.ad._get_update_datetime(
'http://tds0.ifremer.fr/thredds/dodsC/CORIOLIS-ARGO-GDAC-OBS/aoml/5901336/profiles/D5901336_082.nc')
def test_to_odv(self):
df = self.ad.get_float_dataframe(self.good_oga_floats, max_profiles=2)
converters.to_odv(df, 'biofloat_data.txt')
def test_update_cache_false(self):
df = self.ad.get_float_dataframe(self.good_oga_floats, max_profiles=2,
update_cache=False)
self.assertNotEqual(len(df), 0)
def test_remove_df(self):
self.ad._remove_df(self.ad._GLOBAL_META)