How to Convert a Pandas DataFrame to TensorFlow Dataset

Many TensorFlow examples are using TensorFlow Datasets. So the question is, how the heck you can convert your Pandas Data Frame to TensorFlow Datasets? In the example below, we create a dummy Pandas Data Frame and we convert it to a TensorFlow Dataset.

import tensorflow as tf
import pandas as pd

df = pd.DataFrame({'Col1':[1,2,3,4,5],
                   'Col2': ['a','b','c','d','e']})

df

Let’s convert it to a TensorFlow Dataset.

tf_df = tf.data.Dataset.from_tensor_slices(dict(df))
tf_df
 

Let’s print the first 5 rows.

for k in tf_df.take(5):
    print(k)