Representing Data with EntitySets

An EntitySet is a collection of dataframes and the relationships between them. They are useful for preparing raw, structured datasets for feature engineering. While many functions in Featuretools take dataframes and relationships as separate arguments, it is recommended to create an EntitySet, so you can more easily manipulate your data as needed.

The Raw Data

Below we have two tables of data (represented as Pandas DataFrames) related to customer transactions. The first is a merge of transactions, sessions, and customers so that the result looks like something you might see in a log file:

[1]:
import featuretools as ft
data = ft.demo.load_mock_customer()
transactions_df = data["transactions"].merge(data["sessions"]).merge(data["customers"])

transactions_df.sample(10)
[1]:
transaction_id session_id transaction_time product_id amount customer_id device session_start zip_code join_date birthday
264 380 21 2014-01-01 05:14:10 5 57.09 4 desktop 2014-01-01 05:02:15 60091 2011-04-08 20:08:14 2006-08-15
19 244 10 2014-01-01 02:34:55 2 116.95 2 tablet 2014-01-01 02:31:40 13244 2012-04-15 23:31:04 1986-08-18
314 299 6 2014-01-01 01:32:05 4 64.99 1 tablet 2014-01-01 01:23:25 60091 2011-04-17 10:48:33 1994-07-18
290 78 4 2014-01-01 00:54:10 1 37.50 1 mobile 2014-01-01 00:44:25 60091 2011-04-17 10:48:33 1994-07-18
379 457 27 2014-01-01 06:37:35 1 19.16 1 mobile 2014-01-01 06:34:20 60091 2011-04-17 10:48:33 1994-07-18
335 477 9 2014-01-01 02:30:35 3 41.70 1 desktop 2014-01-01 02:15:25 60091 2011-04-17 10:48:33 1994-07-18
293 103 4 2014-01-01 00:57:25 5 20.79 1 mobile 2014-01-01 00:44:25 60091 2011-04-17 10:48:33 1994-07-18
271 390 22 2014-01-01 05:21:45 2 54.83 4 desktop 2014-01-01 05:21:45 60091 2011-04-08 20:08:14 2006-08-15
404 476 29 2014-01-01 07:24:10 4 121.59 1 mobile 2014-01-01 07:10:05 60091 2011-04-17 10:48:33 1994-07-18
179 90 3 2014-01-01 00:35:45 1 75.73 4 mobile 2014-01-01 00:28:10 60091 2011-04-08 20:08:14 2006-08-15

And the second dataframe is a list of products involved in those transactions.

[2]:
products_df = data["products"]
products_df
[2]:
product_id brand
0 1 B
1 2 B
2 3 B
3 4 B
4 5 A

Creating an EntitySet

First, we initialize an EntitySet. If you’d like to give it a name, you can optionally provide an id to the constructor.

[3]:
es = ft.EntitySet(id="customer_data")

Adding dataframes

To get started, we add the transactions dataframe to the EntitySet. In the call to add_dataframe, we specify three important parameters:

  • The index parameter specifies the column that uniquely identifies rows in the dataframe.

  • The time_index parameter tells Featuretools when the data was created.

  • The logical_types parameter indicates that “product_id” should be interpreted as a Categorical column, even though it is just an integer in the underlying data.

[4]:
from woodwork.logical_types import Categorical, PostalCode

es = es.add_dataframe(
    dataframe_name="transactions",
    dataframe=transactions_df,
    index="transaction_id",
    time_index="transaction_time",
    logical_types={
        "product_id": Categorical,
        "zip_code": PostalCode,
    },
)

es
[4]:
Entityset: customer_data
  DataFrames:
    transactions [Rows: 500, Columns: 11]
  Relationships:
    No relationships

You can also use a setter on the EntitySet object to add dataframes

Note

You can also use a setter on the EntitySet object to add dataframes

es["transactions"] = transactions_df

that this will use the default implementation of add_dataframe, notably the following:

  • if the DataFrame does not have Woodwork initialized, the first column will be the index column

  • if the DataFrame does not have Woodwork initialized, all columns will be inferred by Woodwork.

  • if control over the time index column and logical types is needed, Woodwork should be initialized before adding the dataframe.

Note

You can also display your EntitySet structure graphically by calling EntitySet.plot().

This method associates each column in the dataframe to a Woodwork logical type. Each logical type can have an associated standard semantic tag that helps define the column data type. If you don’t specify the logical type for a column, it gets inferred based on the underlying data. The logical types and semantic tags are listed in the schema of the dataframe. For more information on working with logical types and semantic tags, take a look at the Woodwork documention.

[5]:
es["transactions"].ww.schema
[5]:
Logical Type Semantic Tag(s)
Column
transaction_id Integer ['index']
session_id Integer ['numeric']
transaction_time Datetime ['time_index']
product_id Categorical ['category']
amount Double ['numeric']
customer_id Integer ['numeric']
device Categorical ['category']
session_start Datetime []
zip_code PostalCode ['category']
join_date Datetime []
birthday Datetime []

Now, we can do that same thing with our products dataframe.

[6]:
es = es.add_dataframe(
    dataframe_name="products",
    dataframe=products_df,
    index="product_id")

es
[6]:
Entityset: customer_data
  DataFrames:
    transactions [Rows: 500, Columns: 11]
    products [Rows: 5, Columns: 2]
  Relationships:
    No relationships

With two dataframes in our EntitySet, we can add a relationship between them.

Adding a Relationship

We want to relate these two dataframes by the columns called “product_id” in each dataframe. Each product has multiple transactions associated with it, so it is called the parent dataframe, while the transactions dataframe is known as the child dataframe. When specifying relationships, we need four parameters: the parent dataframe name, the parent column name, the child dataframe name, and the child column name. Note that each relationship must denote a one-to-many relationship rather than a relationship which is one-to-one or many-to-many.

[7]:
es = es.add_relationship("products", "product_id", "transactions", "product_id")
es
[7]:
Entityset: customer_data
  DataFrames:
    transactions [Rows: 500, Columns: 11]
    products [Rows: 5, Columns: 2]
  Relationships:
    transactions.product_id -> products.product_id

Now, we see the relationship has been added to our EntitySet.

Creating a dataframe from an existing table

When working with raw data, it is common to have sufficient information to justify the creation of new dataframes. In order to create a new dataframe and relationship for sessions, we “normalize” the transaction dataframe.

[8]:
es = es.normalize_dataframe(
    base_dataframe_name="transactions",
    new_dataframe_name="sessions",
    index="session_id",
    make_time_index="session_start",
    additional_columns=[
        "device",
        "customer_id",
        "zip_code",
        "session_start",
        "join_date",
    ],
)
es
[8]:
Entityset: customer_data
  DataFrames:
    transactions [Rows: 500, Columns: 6]
    products [Rows: 5, Columns: 2]
    sessions [Rows: 35, Columns: 6]
  Relationships:
    transactions.product_id -> products.product_id
    transactions.session_id -> sessions.session_id

Looking at the output above, we see this method did two operations:

  1. It created a new dataframe called “sessions” based on the “session_id” and “session_start” columns in “transactions”

  2. It added a relationship connecting “transactions” and “sessions”

If we look at the schema from the transactions dataframe and the new sessions dataframe, we see two more operations that were performed automatically:

[9]:
es["transactions"].ww.schema
[9]:
Logical Type Semantic Tag(s)
Column
transaction_id Integer ['index']
session_id Integer ['foreign_key', 'numeric']
transaction_time Datetime ['time_index']
product_id Categorical ['foreign_key', 'category']
amount Double ['numeric']
birthday Datetime []
[10]:
es["sessions"].ww.schema
[10]:
Logical Type Semantic Tag(s)
Column
session_id Integer ['index']
device Categorical ['category']
customer_id Integer ['numeric']
zip_code PostalCode ['category']
session_start Datetime ['time_index']
join_date Datetime []
  1. It removed “device”, “customer_id”, “zip_code” and “join_date” from “transactions” and created a new columns in the sessions dataframe. This reduces redundant information as the those properties of a session don’t change between transactions.

  2. It copied and marked “session_start” as a time index column into the new sessions dataframe to indicate the beginning of a session. If the base dataframe has a time index and make_time_index is not set, normalize_dataframe will create a time index for the new dataframe. In this case it would create a new time index called “first_transactions_time” using the time of the first transaction of each session. If we don’t want this time index to be created, we can set make_time_index=False.

If we look at the dataframes, we can see what normalize_dataframe did to the actual data.

[11]:
es["sessions"].head(5)
[11]:
session_id device customer_id zip_code session_start join_date
1 1 desktop 2 13244 2014-01-01 00:00:00 2012-04-15 23:31:04
2 2 mobile 5 60091 2014-01-01 00:17:20 2010-07-17 05:27:50
3 3 mobile 4 60091 2014-01-01 00:28:10 2011-04-08 20:08:14
4 4 mobile 1 60091 2014-01-01 00:44:25 2011-04-17 10:48:33
5 5 mobile 4 60091 2014-01-01 01:11:30 2011-04-08 20:08:14
[12]:
es["transactions"].head(5)
[12]:
transaction_id session_id transaction_time product_id amount birthday
298 298 1 2014-01-01 00:00:00 5 127.64 1986-08-18
2 2 1 2014-01-01 00:01:05 2 109.48 1986-08-18
308 308 1 2014-01-01 00:02:10 3 95.06 1986-08-18
116 116 1 2014-01-01 00:03:15 4 78.92 1986-08-18
371 371 1 2014-01-01 00:04:20 3 31.54 1986-08-18

To finish preparing this dataset, create a “customers” dataframe using the same method call.

[13]:
es = es.normalize_dataframe(
    base_dataframe_name="sessions",
    new_dataframe_name="customers",
    index="customer_id",
    make_time_index="join_date",
    additional_columns=["zip_code", "join_date"],
)

es
[13]:
Entityset: customer_data
  DataFrames:
    transactions [Rows: 500, Columns: 6]
    products [Rows: 5, Columns: 2]
    sessions [Rows: 35, Columns: 4]
    customers [Rows: 5, Columns: 3]
  Relationships:
    transactions.product_id -> products.product_id
    transactions.session_id -> sessions.session_id
    sessions.customer_id -> customers.customer_id

Using the EntitySet

Finally, we are ready to use this EntitySet with any functionality within Featuretools. For example, let’s build a feature matrix for each product in our dataset.

[14]:
feature_matrix, feature_defs = ft.dfs(entityset=es, target_dataframe_name="products")

feature_matrix
[14]:
COUNT(transactions) MAX(transactions.amount) MEAN(transactions.amount) MIN(transactions.amount) SKEW(transactions.amount) STD(transactions.amount) SUM(transactions.amount) MODE(transactions.DAY(birthday)) MODE(transactions.DAY(transaction_time)) MODE(transactions.MONTH(birthday)) ... MODE(transactions.sessions.device) NUM_UNIQUE(transactions.DAY(birthday)) NUM_UNIQUE(transactions.DAY(transaction_time)) NUM_UNIQUE(transactions.MONTH(birthday)) NUM_UNIQUE(transactions.MONTH(transaction_time)) NUM_UNIQUE(transactions.WEEKDAY(birthday)) NUM_UNIQUE(transactions.WEEKDAY(transaction_time)) NUM_UNIQUE(transactions.YEAR(birthday)) NUM_UNIQUE(transactions.YEAR(transaction_time)) NUM_UNIQUE(transactions.sessions.device)
product_id
1 102 149.56 73.429314 6.84 0.125525 42.479989 7489.79 18 1 7 ... desktop 4 1 3 1 4 1 5 1 3
2 92 149.95 76.319891 5.73 0.151934 46.336308 7021.43 18 1 8 ... desktop 4 1 3 1 4 1 5 1 3
3 96 148.31 73.001250 5.89 0.223938 38.871405 7008.12 18 1 8 ... desktop 4 1 3 1 4 1 5 1 3
4 106 146.46 76.311038 5.81 -0.132077 42.492501 8088.97 18 1 7 ... desktop 4 1 3 1 4 1 5 1 3
5 104 149.02 76.264904 5.91 0.098248 42.131902 7931.55 18 1 7 ... mobile 4 1 3 1 4 1 5 1 3

5 rows × 25 columns

As we can see, the features from DFS use the relational structure of our EntitySet. Therefore it is important to think carefully about the dataframes that we create.

Dask and Spark EntitySets

EntitySets can also be created using Dask dataframes or Spark dataframes. For more information refer to Using Dask EntitySets (BETA) and Using Spark EntitySets (BETA).