TensorFlow: Combining Categorical and Continuous Variables
Illia Polosukhin
2017

Hi,

Thanks for your tutorial it is very helpful. I have a question — If I want to perform early stopping, using the same pandas_input_fn, I get the following error:

File “online_tensor1_Hotel.py”, line 306, in <module>
 classifier.fit(input_fn=pandas_input_fn(X_train.loc[0:train_base,],y_train.loc[0:train_base]), steps=2500,monitors=[validation_monitor])
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py”, line 219, in fit
 max_steps=max_steps)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py”, line 507, in _train_model
 max_steps=max_steps)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/graph_actions.py”, line 281, in _supervised_train
 None)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/supervised_session.py”, line 318, in run
 run_metadata=run_metadata)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/recoverable_session.py”, line 54, in run
 run_metadata=run_metadata)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/coordinated_session.py”, line 70, in run
 self._coord.join(self._coordinated_threads_to_join)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/python/training/coordinator.py”, line 383, in join
 six.reraise(*self._exc_info_to_raise)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/coordinated_session.py”, line 66, in run
 return self._sess.run(*args, **kwargs)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/monitored_session.py”, line 107, in run
 induce_stop = monitor.step_end(monitors_step, monitor_outputs)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/monitors.py”, line 396, in step_end
 return self.every_n_step_end(step, output)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/monitors.py”, line 696, in every_n_step_end
 steps=self.eval_steps, metrics=self.metrics, name=self.name)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py”, line 277, in evaluate
 shuffle=False, epochs=1)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/estimators/estimator.py”, line 84, in _get_input_fn
 epochs=epochs)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py”, line 117, in setup_train_data_feeder
 x, y, n_classes, batch_size, shuffle=shuffle, epochs=epochs)
 File “/usr/local/lib/python2.7/site-packages/tensorflow/contrib/learn/python/learn/learn_io/data_feeder.py”, line 249, in __init__
 self._x = check_array(x, dtype=x.dtype)
AttributeError: ‘function’ object has no attribute ‘dtype’

Would you have any suggestions?

Thanks,

Nitika