def _convert_condition(self, condition):
# ref_temp needs to be converted to Temperature type
try:
ref_temp, unit = re.findall(r'([0-9]+).+([CFK])',
condition['ref_temp'])[0]
ref_temp = sigfigs(ref_temp)
condition['ref_temp'] = {'value': ref_temp, 'unit': unit}
except (TypeError, IndexError):
# probably encountered a None, which is fine. We will just leave
# it alone and move to the next one.
pass
# unit needs to be converted to PyNUCOS unit/unit_type
unit_map = {
'% w/w': ('%', 'massfraction'),
'%w/w': ('%', 'massfraction'),
'µg/g': ('µg/g', 'massfraction'),
'mg/g': ('mg/g', 'massfraction'),
'Pa.s': ('Pa.s', 'dynamicviscosity'),
'mPa.s': ('mPa.s', 'dynamicviscosity'),
'Pa': ('Pa', 'pressure'),
'kPa': ('kPa', 'pressure'),
'mN/m or dynes/cm': ('mN/m', 'interfacialtension'),
'g/cm2': ('g/cm^2', 'needleadhesion'),
'g/mL': ('g/mL', 'density'),
'̊C': ('C', 'temperature'),
}
try:
unit, unit_type = unit_map[condition['unit']]
condition['unit'] = unit
condition['unit_type'] = unit_type
except KeyError:
# print(f'unit key "{condition["unit"]}" not found')
pass
def test_dist_cuts(self, samp_ind, cut_index, fraction, temp_f):
samples = ExxonMapper(self.record).sub_samples
cut = samples[samp_ind].distillation_data.cuts[cut_index]
assert cut.fraction.value == fraction
assert isclose(cut.vapor_temp.value,
sigfigs(uc.convert("F", "C", temp_f), 5),
rel_tol=1e-4)
def test_dist_end_point(self, sample_idx, expected):
samples = ExxonMapper(self.record).sub_samples
if expected is None:
assert samples[sample_idx].distillation_data.end_point is None
else:
expected_c = sigfigs(uc.convert("F", "C", expected), 5)
end_point = samples[sample_idx].distillation_data.end_point
assert isclose(end_point.value, expected_c, rel_tol=1e-4)
assert end_point.unit == 'C'
def _get_record_raw_columns(self, name):
"""
Return the columns in the Excel sheet referenced by the name of an
oil.
Note: the Excel sheet columns object has no direct indexing,
only a next(). This is why we are using walk method to get
our indexed columns.
Note: It has been decided that we will only keep 5 significant
digits of any floating point values in the datasheet.
"""
return list(
zip(*[[sigfigs(cell.value, 5) for cell in col]
for i, col in enumerate(self.db_sheet.columns)
if i in self.col_indexes[name]]))
def _get_conditions_columns(self):
"""
The April 2020 update of the Environment Canada datasheet contained
a few extra columns that contained data concerning the testing
conditions for the measurements. They are:
- Unit of Measurement: Instead of annotating the category and/or
field names with unit information, they
put this information into a dedicated column
- Temperature: Instead of annotating temperature information into
the field names, they put this information into a
dedicated column.
- Conditions of analysis: Other significant information concerning
the measurements taken seem to be entered
here. Most significantly, non-newtonian
shear rate for viscosities. This was not
there before.
"""
return list(
zip(*[[sigfigs(strip(cell.value), 5) for cell in col]
for i, col in enumerate(self.db_sheet.columns)
if i in [2, 3, 4]]))
def test_round_sigfigs(val, num_figs, expected):
assert util.sigfigs(val, num_figs) == expected
def set_aggregate_subsample_props(self):
"""
These are properties commonly associated with a sub-sample.
There is a copy of this information inside every measurement,
so we need to reconcile them to determine the identifying
properties of each sub-sample.
Sub-sample properties:
- ests_id: One common value per sub-sample. This could be numeric,
so we force it to be a string.
- weathering_fraction: One value per sub-sample. These values
look like some kind of code that EC uses.
Probably not useful to us.
- weathering_percent: One common value per sub-sample. These
values are mostly a string in the format
'N.N%'. We will convert to a structure
suitable for a Measurement type.
- weathering_method: One common value per sub-sample. This is
information that might be good to save, but
it doesn't fit into the Adios oil model.
"""
first_objs = [
v for v in self.src_values if v['property_id'] == 'Density_0'
]
first_sample_ids = [o['ests_id'] for o in first_objs]
if not self.sample_ids == first_sample_ids:
raise ValueError(f'duplicate sample_ids: {first_sample_ids}')
for idx, o in enumerate(first_objs):
sample_id = str(o['ests_id'])
weathering_percent = o['weathering_percent']
if weathering_percent is None or weathering_percent == 'None':
weathering_percent = None
else:
weathering_percent = {
'value': sigfigs(weathering_percent.rstrip('%'), sig=5),
'unit': '%'
}
if (weathering_percent is not None
and isclose(weathering_percent['value'], 0.0)):
name = 'Fresh Oil Sample'
short_name = 'Fresh Oil'
elif weathering_percent is not None:
name = f'{weathering_percent["value"]}% Evaporated'
short_name = f'{weathering_percent["value"]}% Evaporated'
else:
name = f'{o["weathering_fraction"]}'
short_name = f'{o["weathering_fraction"]}'[:12]
self.deep_set(self.oil_obj,
f'sub_samples.{idx}.metadata.sample_id', sample_id)
self.deep_set(self.oil_obj, f'sub_samples.{idx}.metadata.name',
name)
self.deep_set(self.oil_obj,
f'sub_samples.{idx}.metadata.short_name', short_name)
self.deep_set(self.oil_obj,
f'sub_samples.{idx}.metadata.fraction_weathered',
weathering_percent)