def test_raise_exception_for_observable_with_one_value(observables):
point_constr = {"observables": observables}
params, _ = generate_random_model(point_constr=point_constr)
params = params.drop(index="observable_observable_0_0", level="category")["value"]
with pytest.raises(ValueError, match=r"Observables and exogenous processes"):
_parse_observables({}, params)
def test_parse_observables():
params = pd.read_csv(
io.StringIO(
textwrap.dedent(
"""
category,name,value
observable_fishing_grounds_rich_grounds,probability,0.5
observable_fishing_grounds_poor_grounds,probability,0.5
observable_ability_low_middle,probability,0.5
observable_ability_high,probability,0.5
"""
)
),
index_col=["category", "name"],
)["value"]
optim_paras = _parse_exogenous_processes({}, params)
optim_paras = _parse_observables(optim_paras, params)
expected = {
"fishing_grounds": {
"rich_grounds": pd.Series(data=np.log(0.5), index=["constant"]),
"poor_grounds": pd.Series(data=np.log(0.5), index=["constant"]),
},
"ability": {
"low_middle": pd.Series(data=np.log(0.5), index=["constant"]),
"high": pd.Series(data=np.log(0.5), index=["constant"]),
},
}
for observable, level_dict in optim_paras["observables"].items():
for level in level_dict:
assert optim_paras["observables"][observable][level].equals(
expected[observable][level]
)