-1

I'm trying to display the output of visualize_statistics like below but the results are pretty disappointing:


train_stats = tfdv.generate_statistics_from_tfrecord(data_location='data/tf_records/train.record')
test_stats = tfdv.generate_statistics_from_tfrecord(data_location='data/tf_records/test.record')

schema = tfdv.infer_schema(train_stats)

tfdv.visualize_statistics(lhs_statistics=test_stats, rhs_statistics=train_stats, lhs_name='TEST_DATASET', rhs_name='TRAIN_DATASET')

Result:

<IPython.core.display.HTML object>

How can I improve on this?

I do not want to install Apache Beam

I do not want to install Apache Airflow

I do not want to install Kubeflow

I do not want to display a notebook

I only want to display the HTML.

Chris
  • 11
  • 2

1 Answers1

0

Possibly you could write those statistics to a text file using write_stats_text?

Alex
  • 81
  • 9