Description
pycaret version checks
-
I have checked that this issue has not already been reported here.
-
I have confirmed this bug exists on the latest version of pycaret.
-
I have confirmed this bug exists on the master branch of pycaret (pip install -U git+https://github.com/pycaret/pycaret.git@master).
Issue Description
Hi,
I came across an issue when I set include parameter (in compare_models function) to [] (however, I think it will fail when not model is trained as well). Then, compare_models fails with:
/venv/lib/python3.10/site-packages/pycaret/utils/generic.py:964: in wrapper
return func(*args, **kwargs)
/venv/lib/python3.10/site-packages/pycaret/classification/functional.py:814: in compare_models
return _CURRENT_EXPERIMENT.compare_models(
/venv/lib/python3.10/site-packages/pycaret/classification/oop.py:1180: in compare_models
return_values = super().compare_models(
/venv/lib/python3.10/site-packages/pycaret/internal/pycaret_experiment/supervised_experiment.py:535: in compare_models
return self._parallel_compare_models(parallel, caller_params, turbo=turbo)
/venv/lib/python3.10/site-packages/pycaret/internal/pycaret_experiment/supervised_experiment.py:292: in _parallel_compare_models
return parallel.compare_models(self, params)
/venv/lib/python3.10/site-packages/pycaret/parallel/fugue_backend.py:166: in compare_models
res = pd.concat(cloudpickle.loads(x[0]) for x in outputs)
/venv/lib/python3.10/site-packages/pandas/core/reshape/concat.py:382: in concat
op = _Concatenator(
/venv/lib/python3.10/site-packages/pandas/core/reshape/concat.py:445: in __init__
objs, keys = self._clean_keys_and_objs(objs, keys)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
self = <pandas.core.reshape.concat._Concatenator object at 0x7551f8bedcc0>, objs = <generator object FugueBackend.compare_models.<locals>.<genexpr> at 0x7551f7f2e570>, keys = None
def _clean_keys_and_objs(
self,
objs: Iterable[Series | DataFrame] | Mapping[HashableT, Series | DataFrame],
keys,
) -> tuple[list[Series | DataFrame], Index | None]:
if isinstance(objs, abc.Mapping):
if keys is None:
keys = list(objs.keys())
objs_list = [objs[k] for k in keys]
else:
objs_list = list(objs)
if len(objs_list) == 0:
> raise ValueError("No objects to concatenate")
E ValueError: No objects to concatenate
I found a comment: #3690 (comment)
@Haminh630 compare_models return empty list when model training fails. If you set errors = 'raise' in compare_models you will be able to see what is causing the failure.
Indeed, it returns empty list of models in case of non-parallel mode. So I expect that in parallel mode it should return empty list as well.
Reproducible Example
from pycaret.parallel import FugueBackend
import pycaret.classification as pc
from pycaret.datasets import get_data
pc.setup(
data_func=lambda: get_data("juice", verbose=False),
target="Purchase",
session_id=0,
n_jobs=1,
verbose=False,
html=False
)
fconf = {
"fugue.rpc.server": "fugue.rpc.flask.FlaskRPCServer",
"fugue.rpc.flask_server.host": "localhost",
"fugue.rpc.flask_server.port": "3333",
"fugue.rpc.flask_server.timeout": "2 sec",
}
pc.compare_models(include=[], parallel=FugueBackend("dask", fconf, display_remote=True, batch_size=3, top_only=False))Expected Behavior
When no model is trained then empty list of models is returned.
Actual Results
: No objects to concatenateInstalled Versions
3.2.1