def single_fit(model_version_id: int, location_id: int,
sex_id: int) -> List[_CascadeOperation]:
"""
Create a sequence of tasks to do a single fit both model. Configures
inputs, does a fit fixed, then fit both, then predict and uploads the result.
Will fit the model based on the settings attached to the model version ID.
Parameters
----------
model_version_id
The model version ID.
location_id
The parent location ID to run the model for.
sex_id
The sex ID to run the model for.
Returns
-------
List of CascadeOperations.
"""
t1 = ConfigureInputs(model_version_id=model_version_id)
t2 = Fit(model_version_id=model_version_id,
parent_location_id=location_id,
sex_id=sex_id,
fill=True,
predict=True,
both=True,
save_fit=True,
save_prior=False,
upstream_commands=[t1.command])
t3 = Upload(model_version_id=model_version_id,
fit=True,
upstream_commands=[t2.command])
return [t1, t2, t3]
def single_fit_with_uncertainty(model_version_id: int,
location_id: int,
sex_id: int,
n_sim: int = 100,
n_pool: int = 20) -> List[_CascadeOperation]:
"""
Create a sequence of tasks to do a single fit both model. Configures
inputs, does a fit fixed, then fit both, then predict and uploads the result.
Will fit the model based on the settings attached to the model version ID.
Parameters
----------
model_version_id
The model version ID.
location_id
The parent location ID to run the model for.
sex_id
The sex ID to run the model for.
n_sim
The number of simulations to do, number of draws to make
n_pool
The number of multiprocessing pools to use in creating the draws
Returns
-------
List of CascadeOperations.
"""
t1 = ConfigureInputs(model_version_id=model_version_id)
t2 = Fit(model_version_id=model_version_id,
parent_location_id=location_id,
sex_id=sex_id,
fill=True,
predict=True,
both=True,
save_fit=True,
upstream_commands=[t1.command])
t3 = Sample(model_version_id=model_version_id,
parent_location_id=location_id,
sex_id=sex_id,
n_sim=n_sim,
n_pool=n_pool,
fit_type='both',
upstream_commands=[t2.command],
executor_parameters={'num_cores': n_pool},
asymptotic=True)
t4 = Predict(model_version_id=model_version_id,
parent_location_id=location_id,
sex_id=sex_id,
save_final=True,
prior_grid=False,
sample=True,
upstream_commands=[t3.command])
t5 = Upload(model_version_id=model_version_id,
fit=True,
final=True,
upstream_commands=[t4.command])
return [t1, t2, t3, t4, t5]
def test_configure_inputs():
obj = ConfigureInputs(model_version_id=0)
assert obj.command == (f'configure_inputs --model-version-id 0 '
f'--make --configure')
assert obj.template_kwargs['model_version_id'] == '--model-version-id 0'
assert obj.template_kwargs['make'] == '--make'
assert obj.template_kwargs['configure'] == '--configure'
assert obj.template_kwargs['log_level'] == '--log-level info'
assert obj.template_kwargs['json_file'] == ''
assert obj.template_kwargs['test_dir'] == ''
assert obj.name == 'dmat_configure_inputs_0'
def root_fit(model_version_id: int,
location_id: int,
sex_id: int,
child_locations: List[int],
child_sexes: List[int],
skip_configure: bool = False,
mulcov_stats: bool = True,
n_sim: int = 10,
n_pool: int = 10) -> List[_CascadeOperation]:
"""
Create a sequence of tasks to do a top-level prior fit.
Does a fit fixed, then fit both, then creates posteriors
that can be used as priors later on. Saves its fit to be uploaded.
Parameters
----------
model_version_id
The model version ID.
location_id
The parent location ID to run the model for.
sex_id
The sex ID to run the model for.
child_locations
The children to fill the avgint table with
child_sexes
The sexes to predict for.
skip_configure
Don't run a task to configure the inputs. Only do this if it has already happened.
mulcov_stats
Compute mulcov statistics at this level
n_sim
n_pool
Returns
-------
List of CascadeOperations.
"""
tasks = []
if not skip_configure:
t1 = ConfigureInputs(model_version_id=model_version_id)
upstream = [t1.command]
tasks.append(t1)
else:
upstream = None
t2 = Fit(model_version_id=model_version_id,
parent_location_id=location_id,
sex_id=sex_id,
fill=True,
both=False,
predict=True,
upstream_commands=upstream,
save_fit=True)
tasks.append(t2)
t3 = Predict(model_version_id=model_version_id,
parent_location_id=location_id,
sex_id=sex_id,
child_locations=child_locations,
child_sexes=child_sexes,
sample=False,
upstream_commands=[t2.command])
tasks.append(t3)
if mulcov_stats:
t4 = Sample(model_version_id=model_version_id,
parent_location_id=location_id,
sex_id=sex_id,
n_sim=n_sim,
n_pool=n_pool,
fit_type='fixed',
asymptotic=True,
upstream_commands=[t3.command],
executor_parameters={'num_cores': n_pool})
tasks.append(t4)
t5 = MulcovStatistics(model_version_id=model_version_id,
locations=[location_id],
sexes=[sex_id],
sample=True,
mean=True,
std=True,
quantile=[0.025, 0.975],
upstream_commands=[t4.command])
tasks.append(t5)
return tasks
def test_configure_inputs():
obj = ConfigureInputs(model_version_id=0)
assert obj.command == (f'configure_inputs --model-version-id 0 '
f'--make --configure')