1

My target feature(frame strength) is not an unique value. I have train and test dataset. How can I approach to use Ft? My datasets feature are temperature, hive size, some percentile values, some entropy, different Pixel, Frame size etc..

I tried to DFS but as I don't have any unique ID, couldn't implement.

HMI
  • 11
  • 1

1 Answers1

1

You do not need an unique identifier in your dataset to use Featuretools. You can tell Featuretools to make an index column.

You can set make_index to True in your call to add_dataframe to create a new index on that data - make_index creates a unique index for each row by just looking at what number the row is, in relation to all the other rows. The name of the new index is controlled from the index parameter.

product_df = pd.DataFrame({"product": [1, 2, 3, 4, 4], 
                           "rating": [3.5, 4.0, 4.5, 1.5, 5.0]})
product_df
es = ft.EntitySet(id="product_data")
es = es.add_dataframe(dataframe_name="products", 
                      dataframe=product_df, 
                      make_index=True,
                      index="id")
es["products"]
  • If you look at the products table in the EntitySet, you will see the newly created index column.
Gaurav
  • 358
  • 5
  • 18