ValueError: The number of quantiles cannot be greater than the number of samples used

[mgcyung]

Training worked for 100000 rows on TabPFNClassifier, while predicting didn't

train_size=200000
test_size=10000
clf = TabPFNClassifier(memory_saving_mode=16, fit_mode='low_memory', ignore_pretraining_limits=True)
clf.fit(X[:train_size,:].astype(float), y[:train_size])
predictions = clf.predict(X[-test_size:,:].astype(float))

And the error said

ValueError: The number of quantiles cannot be greater than the number of samples used. Got 20000 quantiles and 10000 samples.

[noahho]

Hi! 200,000 samples will be hard to fit memory-wise. I'm not sure where your error here comes from, can you provide a fuller trace please?

Best!

[mgcyung]

Here is the trace

Traceback (most recent call last):
  File "/home/test/tabpfn_test.py", line 73, in <module>
    predictions = clf.predict(X[-test_size:,:].astype(float))
  File "/home/test/.local/lib/python3.10/site-packages/tabpfn/classifier.py", line 512, in predict
    proba = self.predict_proba(X)
  File "/home/test/.local/lib/python3.10/site-packages/tabpfn/classifier.py", line 533, in predict_proba
    for output, config in self.executor_.iter_outputs(
  File "/home/test/.local/lib/python3.10/site-packages/tabpfn/inference.py", line 163, in iter_outputs
    for config, preprocessor, X_train, y_train, cat_ix in itr:
  File "/home/test/.local/lib/python3.10/site-packages/tabpfn/preprocessing.py", line 632, in fit_preprocessing
    yield from executor(  # type: ignore
  File "/home/test/.local/lib/python3.10/site-packages/joblib/parallel.py", line 1792, in _get_sequential_output
    res = func(*args, **kwargs)
  File "/home/test/.local/lib/python3.10/site-packages/tabpfn/preprocessing.py", line 539, in fit_preprocessing_one
    res = preprocessor.fit_transform(X_train, cat_ix)
  File "/home/test/.local/lib/python3.10/site-packages/tabpfn/model/preprocessing.py", line 397, in fit_transform
    X, categorical_features = step.fit_transform(X, categorical_features)
  File "/home/test/.local/lib/python3.10/site-packages/tabpfn/model/preprocessing.py", line 987, in fit_transform
    Xt = transformer.fit_transform(X[:, self.subsampled_features_])
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/base.py", line 1152, in wrapper
    return fit_method(estimator, *args, **kwargs)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/pipeline.py", line 471, in fit_transform
    Xt = self._fit(X, y, **fit_params_steps)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/pipeline.py", line 377, in _fit
    X, fitted_transformer = fit_transform_one_cached(
  File "/home/test/.local/lib/python3.10/site-packages/joblib/memory.py", line 353, in __call__
    return self.func(*args, **kwargs)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/pipeline.py", line 957, in _fit_transform_one
    res = transformer.fit_transform(X, y, **fit_params)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/utils/_set_output.py", line 157, in wrapped
    data_to_wrap = f(self, X, *args, **kwargs)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/base.py", line 1152, in wrapper
    return fit_method(estimator, *args, **kwargs)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py", line 754, in fit_transform
    result = self._fit_transform(X, y, _fit_transform_one)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/compose/_column_transformer.py", line 681, in _fit_transform
    return Parallel(n_jobs=self.n_jobs)(
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/utils/parallel.py", line 65, in __call__
    return super().__call__(iterable_with_config)
  File "/home/test/.local/lib/python3.10/site-packages/joblib/parallel.py", line 1863, in __call__
    return output if self.return_generator else list(output)
  File "/home/test/.local/lib/python3.10/site-packages/joblib/parallel.py", line 1792, in _get_sequential_output
    res = func(*args, **kwargs)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/utils/parallel.py", line 127, in __call__
    return self.function(*args, **kwargs)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/pipeline.py", line 957, in _fit_transform_one
    res = transformer.fit_transform(X, y, **fit_params)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/utils/_set_output.py", line 157, in wrapped
    data_to_wrap = f(self, X, *args, **kwargs)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/base.py", line 916, in fit_transform
    return self.fit(X, **fit_params).transform(X)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/base.py", line 1152, in wrapper
    return fit_method(estimator, *args, **kwargs)
  File "/home/test/.local/lib/python3.10/site-packages/sklearn/preprocessing/_data.py", line 2653, in fit
    raise ValueError(
ValueError: The number of quantiles cannot be greater than the number of samples used. Got 20000 quantiles and 10000 samples.

[ChenJin1110]

Hi, noahho!

I encountered the same error when using the training set of more than 10,000 rows. My x_train.shape=(273835, 85).

Is there any solution?

ValueError: The number of quantiles cannot be greater than the number of samples used. Got 54767 quantiles and 10000 samples.

[ljubomirj]

Similar for me - sizes

X: (4542148, 15) Y: (4542148, 1)

running

model = TabPFNRegressor(ignore_pretraining_limits = True)
model.fit(X, Y)

got error

...........................................................................................
File "/home/ljubomir/python3-venv/torch/lib/python3.10/site-packages/joblib/parallel.py", line 1847, in _get_sequential_output
res = func(*args, **kwargs)
File "/home/ljubomir/python3-venv/torch/lib/python3.10/site-packages/sklearn/utils/parallel.py", line 136, in call
return self.function(*args, **kwargs)
File "/home/ljubomir/python3-venv/torch/lib/python3.10/site-packages/sklearn/pipeline.py", line 1310, in _fit_transform_one
res = transformer.fit_transform(X, y, **params.get("fit_transform", {}))
File "/home/ljubomir/python3-venv/torch/lib/python3.10/site-packages/sklearn/utils/_set_output.py", line 316, in wrapped
data_to_wrap = f(self, X, *args, **kwargs)
File "/home/ljubomir/python3-venv/torch/lib/python3.10/site-packages/sklearn/base.py", line 1098, in fit_transform
return self.fit(X, **fit_params).transform(X)
File "/home/ljubomir/python3-venv/torch/lib/python3.10/site-packages/sklearn/base.py", line 1473, in wrapper
return fit_method(estimator, *args, **kwargs)
File "/home/ljubomir/python3-venv/torch/lib/python3.10/site-packages/sklearn/preprocessing/_data.py", line 2775, in fit
raise ValueError(
ValueError: The number of quantiles cannot be greater than the number of samples used. Got 908429 quantiles and 10000 samples.

[1511878618]

Same errors:

ValueError: The number of quantiles cannot be greater than the number of samples used. Got 45374 quantiles and 10000 samples.

[LeoGrin]

Duplicate of #169