readme examples does not work because Inferencer.pre_fit() got an unexpected keyword argument (offset)

[timothee-clemessy]

Trying to run the examples from the readme :

from yaglm.toy_data import sample_sparse_lin_reg

from yaglm.GlmTuned import GlmCV, GlmTrainMetric

from yaglm.config.loss import Huber
from yaglm.config.penalty import Lasso, GroupLasso
from yaglm.config.flavor import Adaptive, NonConvex

from yaglm.metrics.info_criteria import InfoCriteria
from yaglm.infer.Inferencer import Inferencer
from yaglm.infer.lin_reg_noise_var import ViaRidge


# sample sparse linear regression data
X, y, _ = sample_sparse_lin_reg(n_samples=100, n_features=10)

# fit a lasso penalty tuned via cross-validation with the 1se rule
GlmCV(loss='lin_reg',
      penalty=Lasso(),  # specify penalty with config object
      select_rule='1se'
      ).fit(X, y)

# fit an adaptive lasso tuned via cross-validation
# initialized with a lasso tuned with cross-validation
GlmCV(loss='lin_reg',
      penalty=Lasso(flavor=Adaptive()),
      initializer='default'
      ).fit(X, y)

# fit an adaptive lasso and tuned via EBIC
# estimate the noise variance via a ridge-regression method
GlmTrainMetric(loss='lin_reg',
               penalty=Lasso(flavor=Adaptive()),

               inferencer=Inferencer(scale=ViaRidge()),  # noise variance estimator
               scorer=InfoCriteria(crit='ebic')  # Info criteria
               ).fit(X, y)

raises the following errors :

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In [11], line 37
     25 GlmCV(loss='lin_reg',
     26       penalty=Lasso(flavor=Adaptive()),
     27       initializer='default'
     28       ).fit(X, y)
     30 # fit an adaptive lasso and tuned via EBIC
     31 # estimate the noise variance via a ridge-regression method
     32 GlmTrainMetric(loss='lin_reg',
     33                penalty=Lasso(flavor=Adaptive()),
     34 
     35                inferencer=Inferencer(scale=ViaRidge()),  # noise variance estimator
     36                scorer=InfoCriteria(crit='ebic')  # Info criteria
---> 37                ).fit(X, y)

File ~/opt/anaconda3/envs/automated_reliability_tests/lib/python3.10/site-packages/yaglm-_0.3.3_-py3.10.egg/yaglm/GlmTuned.py:341, in GlmTrainMetric.fit(self, X, y, sample_weight, offsets)
    334 start_time = time()
    336 ##############################################
    337 # setup, preprocess, and prefitting routines #
    338 ##############################################
    339 pro_data, raw_data, pre_pro_out, \
    340     configs, solver, init_data, inferencer = \
--> 341     self.setup_and_prefit(X=X, y=y,
    342                           sample_weight=sample_weight,
    343                           offsets=offsets)
    345 # store inferencer
    346 self.inferencer_ = inferencer

File ~/opt/anaconda3/envs/automated_reliability_tests/lib/python3.10/site-packages/yaglm-_0.3.3_-py3.10.egg/yaglm/base.py:429, in BaseGlm.setup_and_prefit(self, X, y, sample_weight, offsets)
    425 raw_data = {'X': X, 'y': y,
    426             'sample_weight': sample_weight, 'offsets': offsets}
    428 # run any prefitting inference
--> 429 inferencer = self.run_prefit_inference(**raw_data)
    431 # preproceess X, y
    432 pro_data, pre_pro_out = self.preprocess(**raw_data, copy=True)

File ~/opt/anaconda3/envs/automated_reliability_tests/lib/python3.10/site-packages/yaglm-_0.3.3_-py3.10.egg/yaglm/base.py:718, in BaseGlm.run_prefit_inference(self, X, y, sample_weight, offsets)
    715 if self.inferencer is not None:
    716     # TODO: do we want to do a copy here?
    717     inferencer = deepcopy(self.inferencer)
--> 718     inferencer.pre_fit(estimator=self, X=X, y=y,
    719                        sample_weight=sample_weight,
    720                        offsets=offsets)
    721     return inferencer
    723 else:

TypeError: Inferencer.pre_fit() got an unexpected keyword argument 'offset

Indeed, l718 in base.py a call to pre_fit of an inferencer is passed using an offset argument :

            inferencer.pre_fit(estimator=self, X=X, y=y,
                               sample_weight=sample_weight,
                               offsets=offsets)

while inferencer pre_fit methods does not accept offsets parameter yet according to l45 in Inferencer.py

def pre_fit(self, estimator, X, y, sample_weight=None):

I see that a commit adding offset have been merged recently so I guess it is related.

I hope it helps,

Thanks a lot for the package and I hope the project goes on !

Best regards,

Timothée

[t-silvers]

hi @idc9 and @SteveDiamond, any progress on this issue? still hitting this error too on 0.3.3

[idc9]

I'm sorry I totally overlooked this iseue -- will get back soon!