def get_confint(self, _id):
confint = mongo.data(_id)
if not len(confint):
return []
else:
confint = np.array(confint)
return confint - confint.mean(1)[:, None]
def get_confint(self, _id):
confint = mongo.data(_id)
if not len(confint):
return []
else:
confint = np.array(confint)
return confint - confint.mean(1)[:,None]
def generate(self):
"""Generate the view of differenced production."""
for _id in self._ids:
data = mongo.data(_id)
degree = mongo.settings(_id)['degree']
self.plot(data, degree)
plt.clf()
def generate(self):
"""Generate the view of smoothed production."""
for _id in self._ids:
data = mongo.data(_id)
step = mongo.settings(_id)['step']
self.plot(data, step)
plt.clf()
def generate(self):
"""Generate the view of differenced production."""
for _id in self._ids:
data = mongo.data(_id)
degree = mongo.settings(_id)['degree']
self.plot(data, degree)
plt.clf()
def generate(self):
"""Generate the view of smoothed production."""
for _id in self._ids:
data = mongo.data(_id)
step = mongo.settings(_id)['step']
self.plot(data, step)
plt.clf()
def generate(self):
"""Generate the view of the correlogram."""
for _id in self._ids:
data = mongo.data(_id[LABELS['values']])
confint = self.get_confint(_id[LABELS['confint']])
stg = mongo.settings(mongo.parents(_id[LABELS['values']])[0])
title = ' - '.join(
['{} = {}'.format(k, v) for k, v in stg.items()])
self.plot(data, confint, title)
def generate(self):
"""Generate the view of the correlogram."""
for _id in self._ids:
data = mongo.data(_id[LABELS['values']])
confint = self.get_confint(_id[LABELS['confint']])
stg = mongo.settings(
mongo.parents(_id[LABELS['values']])[0]
)
title = ' - '.join(['{} = {}'.format(k, v) for k, v in stg.items()])
self.plot(data, confint, title)
def wrapper(settings, sources):
"""Prepare the data for the analysis then run it and save it if
necessary.
"""
parents = sorted([_id for name, _id in sources.items()])
created_ids = {}
data = {name: mongo.data(_id) for name, _id in sources.items()}
need_analysis = not all([
mongo.is_analysis(type_, label, parents, stg)
for label, stg in settings.items()
])
if need_analysis or FORCE_CACHE:
data = func(data, settings)
for label, stg in settings.items():
saved = mongo.is_analysis(type_, label, parents, stg)
if not saved:
_id = mongo.db.analysis.insert_one({
'type': type_,
'label': label,
'parents': parents,
'settings': stg,
'data': data[label],
}).inserted_id
else:
_id = mongo.db.analysis.find_one({
'type': type_,
'label': label,
'parents': parents,
'settings': stg,
})['_id']
if FORCE_CACHE:
mongo.db.analysis.update({
'_id': _id,
}, {
'$set': {
'data': data[label]
}
}, upsert=False)
created_ids[label] = _id
return created_ids
def main():
"""Launch the analysis."""
for cow in mongo.cows():
# Crude
dta = mongo.identity(cow, LABELS['prods'])
_id = dta['_id']
views.Crude(_id).render()
acf_ids = analysis.acf({
LABELS['values']: {},
LABELS['confint']: {'alpha': 0.05},
}, {
'data': _id,
})
views.ACF('production', 'ACF', [acf_ids]).render()
pacf_ids = analysis.pacf({
LABELS['values']: {},
LABELS['confint']: {'alpha': 0.05},
}, {
'data': _id,
})
views.PACF('production', 'PACF', [pacf_ids]).render()
# Smoothing
smooth_ids = []
steps = [7, 30]
for step in steps:
id_ = smoothing(_id, step)[LABELS['values']]
smooth_ids.append(id_)
# R: TODO
if step == 7:
sdta = mongo.data(id_)
views.Smoothing('production', 'Smoothed production',
smooth_ids).render()
# Differencing
diff_ids = []
acf_ids = []
pacf_ids = []
degrees = [1]
for degree in degrees:
id_ = differencing(_id, degree)[LABELS['values']]
diff_ids.append(id_)
acf_ids.append(analysis.acf({
LABELS['values']: {},
LABELS['confint']: {'alpha': 0.05},
}, {
'data': id_,
}))
pacf_ids.append(analysis.pacf({
LABELS['values']: {},
LABELS['confint']: {'alpha': 0.05},
}, {
'data': id_,
}))
views.Differencing('production', 'Differenced production',
diff_ids).render()
views.ACF(join('production', 'differenced'), 'ACF', acf_ids).render()
views.PACF(join('production', 'differenced'), 'PACF', pacf_ids).render()
def main():
"""Launch the analysis."""
for cow in mongo.cows():
# Crude
dta = mongo.identity(cow, LABELS['prods'])
_id = dta['_id']
views.Crude(_id).render()
acf_ids = analysis.acf(
{
LABELS['values']: {},
LABELS['confint']: {
'alpha': 0.05
},
}, {
'data': _id,
})
views.ACF('production', 'ACF', [acf_ids]).render()
pacf_ids = analysis.pacf(
{
LABELS['values']: {},
LABELS['confint']: {
'alpha': 0.05
},
}, {
'data': _id,
})
views.PACF('production', 'PACF', [pacf_ids]).render()
# Smoothing
smooth_ids = []
steps = [7, 30]
for step in steps:
id_ = smoothing(_id, step)[LABELS['values']]
smooth_ids.append(id_)
# R: TODO
if step == 7:
sdta = mongo.data(id_)
views.Smoothing('production', 'Smoothed production',
smooth_ids).render()
# Differencing
diff_ids = []
acf_ids = []
pacf_ids = []
degrees = [1]
for degree in degrees:
id_ = differencing(_id, degree)[LABELS['values']]
diff_ids.append(id_)
acf_ids.append(
analysis.acf(
{
LABELS['values']: {},
LABELS['confint']: {
'alpha': 0.05
},
}, {
'data': id_,
}))
pacf_ids.append(
analysis.pacf(
{
LABELS['values']: {},
LABELS['confint']: {
'alpha': 0.05
},
}, {
'data': id_,
}))
views.Differencing('production', 'Differenced production',
diff_ids).render()
views.ACF(join('production', 'differenced'), 'ACF', acf_ids).render()
views.PACF(join('production', 'differenced'), 'PACF',
pacf_ids).render()
