Frequently Asked Questions#

Here we are attempting to answer some commonly asked questions that appear on Github, and Stack Overflow.

[1]:
import featuretools as ft
import pandas as pd
import numpy as np
import woodwork as ww

EntitySet#

How do I get a list of column names and types in an EntitySet?#

After you create your EntitySet, you may wish to view the column names. An EntitySet contains multiple DataFrames, one for each table in the EntitySet.

[2]:
es = ft.demo.load_mock_customer(return_entityset=True)
es
[2]:
Entityset: transactions
  DataFrames:
    transactions [Rows: 500, Columns: 6]
    products [Rows: 5, Columns: 3]
    sessions [Rows: 35, Columns: 5]
    customers [Rows: 5, Columns: 5]
  Relationships:
    transactions.product_id -> products.product_id
    transactions.session_id -> sessions.session_id
    sessions.customer_id -> customers.customer_id

If you want to view the underlying Dataframe, you can do the following:

[3]:
es["transactions"].head()
[3]:
transaction_id session_id transaction_time product_id amount _ft_last_time
298 298 1 2014-01-01 00:00:00 5 127.64 2014-01-01 00:00:00
2 2 1 2014-01-01 00:01:05 2 109.48 2014-01-01 00:01:05
308 308 1 2014-01-01 00:02:10 3 95.06 2014-01-01 00:02:10
116 116 1 2014-01-01 00:03:15 4 78.92 2014-01-01 00:03:15
371 371 1 2014-01-01 00:04:20 3 31.54 2014-01-01 00:04:20

If you want view the columns and types for the “transactions” DataFrame, you can do the following:

[4]:
es["transactions"].ww
[4]:
Physical Type Logical Type Semantic Tag(s)
Column
transaction_id int64 Integer ['index']
session_id int64 Integer ['foreign_key', 'numeric']
transaction_time datetime64[ns] Datetime ['time_index']
product_id category Categorical ['foreign_key', 'category']
amount float64 Double ['numeric']
_ft_last_time datetime64[ns] Datetime ['last_time_index']

What is the difference between copy_columns and additional_columns?#

The function normalize_dataframe creates a new DataFrame and a relationship from unique values of an existing DataFrame. It takes 2 similar arguments:

  • additional_columns removes columns from the base DataFrame and moves them to the new DataFrame.

  • copy_columns keeps the given columns in the base DataFrame, but also copies them to the new DataFrame.

[5]:
data = ft.demo.load_mock_customer()
transactions_df = data["transactions"].merge(data["sessions"]).merge(data["customers"])
products_df = data["products"]

es = ft.EntitySet(id="customer_data")
es = es.add_dataframe(
    dataframe_name="transactions",
    dataframe=transactions_df,
    index="transaction_id",
    time_index="transaction_time",
)

es = es.add_dataframe(
    dataframe_name="products", dataframe=products_df, index="product_id"
)

es = es.add_relationship("products", "product_id", "transactions", "product_id")

Before we normalize to create a new DataFrame, let’s look at the base DataFrame

[6]:
es["transactions"].head()
[6]:
transaction_id session_id transaction_time product_id amount customer_id device session_start zip_code join_date birthday
298 298 1 2014-01-01 00:00:00 5 127.64 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18
2 2 1 2014-01-01 00:01:05 2 109.48 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18
308 308 1 2014-01-01 00:02:10 3 95.06 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18
116 116 1 2014-01-01 00:03:15 4 78.92 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18
371 371 1 2014-01-01 00:04:20 3 31.54 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18

Notice the columns session_id, session_start, join_date, device, customer_id, and zip_code.

[7]:
es = es.normalize_dataframe(
    base_dataframe_name="transactions",
    new_dataframe_name="sessions",
    index="session_id",
    make_time_index="session_start",
    additional_columns=["join_date"],
    copy_columns=["device", "customer_id", "zip_code", "session_start"],
)

Above, we normalized the columns to create a new DataFrame.

  • For additional_columns, the following column ['join_date] will be removed from the transactions DataFrame, and moved to the new sessions DataFrame.

  • For copy_columns, the following columns ['device', 'customer_id', 'zip_code','session_start'] will be copied from the transactions DataFrame to the new sessions DataFrame.

Let’s see this in the actual EntitySet.

[8]:
es["transactions"].head()
[8]:
transaction_id session_id transaction_time product_id amount customer_id device session_start zip_code birthday
298 298 1 2014-01-01 00:00:00 5 127.64 2 desktop 2014-01-01 13244 1986-08-18
2 2 1 2014-01-01 00:01:05 2 109.48 2 desktop 2014-01-01 13244 1986-08-18
308 308 1 2014-01-01 00:02:10 3 95.06 2 desktop 2014-01-01 13244 1986-08-18
116 116 1 2014-01-01 00:03:15 4 78.92 2 desktop 2014-01-01 13244 1986-08-18
371 371 1 2014-01-01 00:04:20 3 31.54 2 desktop 2014-01-01 13244 1986-08-18

Notice above how ['device', 'customer_id', 'zip_code','session_start'] are still in the transactions DataFrame, while ['join_date'] is not. But, they have all been moved to the sessions DataFrame, as seen below.

[9]:
es["sessions"].head()
[9]:
session_id join_date device customer_id zip_code session_start
1 1 2012-04-15 23:31:04 desktop 2 13244 2014-01-01 00:00:00
2 2 2010-07-17 05:27:50 mobile 5 60091 2014-01-01 00:17:20
3 3 2011-04-08 20:08:14 mobile 4 60091 2014-01-01 00:28:10
4 4 2011-04-17 10:48:33 mobile 1 60091 2014-01-01 00:44:25
5 5 2011-04-08 20:08:14 mobile 4 60091 2014-01-01 01:11:30

Why did my columns get new semantic tags?#

During the creation of your EntitySet, you might be wondering why the semantic tags in your columns change.

[10]:
data = ft.demo.load_mock_customer()
transactions_df = data["transactions"].merge(data["sessions"]).merge(data["customers"])
products_df = data["products"]

es = ft.EntitySet(id="customer_data")
es = es.add_dataframe(
    dataframe_name="transactions",
    dataframe=transactions_df,
    index="transaction_id",
    time_index="transaction_time",
)
es.plot()
[10]:
../_images/resources_frequently_asked_questions_20_0.svg

If a column contains semantic tags, they will appear on the right side of a semicolon in the plot above. Notice how session_id and session_start do not have any semantic tags currently associated to them.

Now, let’s normalize the transactions DataFrame to create a new DataFrame.

[11]:
es = es.normalize_dataframe(
    base_dataframe_name="transactions",
    new_dataframe_name="sessions",
    index="session_id",
    make_time_index="session_start",
    additional_columns=["session_start"],
)
es.plot()
[11]:
../_images/resources_frequently_asked_questions_22_0.svg

The session_id now has the sematic tag foreign_key in the transactions DataFrame, and index in the new DataFrame, sessions. This is the case because when we normalize the DataFrame, we create a new relationship between the transactions and sessions. There is a one to many relationship between the parent DataFrame, sessions, and child DataFrame, transactions.

Therefore, session_id has the semantic tag foreign_key in transactions because it represents an index in another DataFrame. There would be a similar effect if we added another DataFrame using add_dataframe and add_relationship.

In addition, when we created the new DataFrame, we set session_start as the time_index. This added the semantic tag time_index to the session_start column in the new sessions DataFrame because it now represents a time_index.

How do I update a column’s description or metadata?#

You can directly update the description or metadata attributes of the column schema. However, you must specifically use the column schema returned by DataFrame.ww.columns['col_name'], not DataFrame.ww['col_name'].ww.schema. The column schema from DataFrame.ww.columns['col_name'] is still associated with the EntitySet and propagates any attribute updates, whereas the other does not. As an example, this is how you can update a column’s description or metadata:

column_schema = df.ww.columns['col_name']
column_schema.description = 'my description'
column_schema.metadata.update(key='value')

How do I combine two or more interesting values?#

You might want to create features that are conditioned on multiple values before they are calculated. This would require the use of interesting_values. However, since we are trying to create the feature with multiple conditions, we will need to modify the Dataframe before we create the EntitySet.

Let’s look at how you might accomplish this.

First, let’s create our Dataframes.

[12]:
data = ft.demo.load_mock_customer()
transactions_df = data["transactions"].merge(data["sessions"]).merge(data["customers"])
products_df = data["products"]
[13]:
transactions_df.head()
[13]:
transaction_id session_id transaction_time product_id amount customer_id device session_start zip_code join_date birthday
0 298 1 2014-01-01 00:00:00 5 127.64 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18
1 2 1 2014-01-01 00:01:05 2 109.48 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18
2 308 1 2014-01-01 00:02:10 3 95.06 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18
3 116 1 2014-01-01 00:03:15 4 78.92 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18
4 371 1 2014-01-01 00:04:20 3 31.54 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18
[14]:
products_df.head()
[14]:
product_id brand
0 1 B
1 2 B
2 3 B
3 4 B
4 5 A

Now, let’s modify our transactions Dataframe to create the additional column that represents multiple conditions for our feature.

[15]:
transactions_df["product_id_device"] = (
    transactions_df["product_id"].astype(str) + " and " + transactions_df["device"]
)

Here, we created a new column called product_id_device, which just combines the product_id column, and the device column.

Now let’s create our EntitySet.

[16]:
es = ft.EntitySet(id="customer_data")
es = es.add_dataframe(
    dataframe_name="transactions",
    dataframe=transactions_df,
    index="transaction_id",
    time_index="transaction_time",
    logical_types={
        "product_id": ww.logical_types.Categorical,
        "product_id_device": ww.logical_types.Categorical,
        "zip_code": ww.logical_types.PostalCode,
    },
)

es = es.add_dataframe(
    dataframe_name="products", dataframe=products_df, index="product_id"
)

es = es.normalize_dataframe(
    base_dataframe_name="transactions",
    new_dataframe_name="sessions",
    index="session_id",
    additional_columns=["device", "product_id_device", "customer_id"],
)

es = es.normalize_dataframe(
    base_dataframe_name="sessions", new_dataframe_name="customers", index="customer_id"
)
es
[16]:
Entityset: customer_data
  DataFrames:
    transactions [Rows: 500, Columns: 9]
    products [Rows: 5, Columns: 2]
    sessions [Rows: 35, Columns: 5]
    customers [Rows: 5, Columns: 2]
  Relationships:
    transactions.session_id -> sessions.session_id
    sessions.customer_id -> customers.customer_id

Now, we are ready to add our interesting values.

First, let’s view our options for what the interesting values could be.

[17]:
interesting_values = transactions_df["product_id_device"].unique().tolist()
interesting_values
[17]:
['5 and desktop',
 '2 and desktop',
 '3 and desktop',
 '4 and desktop',
 '1 and desktop',
 '4 and mobile',
 '5 and mobile',
 '1 and mobile',
 '3 and mobile',
 '2 and mobile',
 '4 and tablet',
 '3 and tablet',
 '2 and tablet',
 '1 and tablet',
 '5 and tablet']

If you wanted to, you could pick a subset of these, and the where features created would only use those conditions. In our example, we will use all the possible interesting values.

Here, we set all of these values as our interesting values for this specific DataFrame and column. If we wanted to, we could make interesting values in the same way for more than one column, but we will just stick with this one for this example.

[18]:
values = {"product_id_device": interesting_values}
es.add_interesting_values(dataframe_name="sessions", values=values)

Now we can run DFS.

[19]:
feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    agg_primitives=["count"],
    where_primitives=["count"],
    trans_primitives=[],
)
feature_matrix.head()
[19]:
COUNT(sessions) COUNT(transactions) COUNT(sessions WHERE product_id_device = 1 and tablet) COUNT(sessions WHERE product_id_device = 2 and tablet) COUNT(sessions WHERE product_id_device = 4 and desktop) COUNT(sessions WHERE product_id_device = 1 and desktop) COUNT(sessions WHERE product_id_device = 5 and mobile) COUNT(sessions WHERE product_id_device = 5 and tablet) COUNT(sessions WHERE product_id_device = 4 and tablet) COUNT(sessions WHERE product_id_device = 3 and tablet) ... COUNT(transactions WHERE sessions.product_id_device = 2 and tablet) COUNT(transactions WHERE sessions.product_id_device = 4 and mobile) COUNT(transactions WHERE sessions.product_id_device = 5 and tablet) COUNT(transactions WHERE sessions.product_id_device = 5 and desktop) COUNT(transactions WHERE sessions.product_id_device = 5 and mobile) COUNT(transactions WHERE sessions.product_id_device = 4 and tablet) COUNT(transactions WHERE sessions.product_id_device = 2 and desktop) COUNT(transactions WHERE sessions.product_id_device = 4 and desktop) COUNT(transactions WHERE sessions.product_id_device = 3 and tablet) COUNT(transactions WHERE sessions.product_id_device = 1 and tablet)
customer_id
2 7 93 1 0 1 1 0 1 0 0 ... 0 18 13 16 0 0 0 10 0 15
5 6 79 0 0 1 0 0 0 1 0 ... 0 10 0 15 0 14 0 14 0 0
4 8 109 0 0 1 0 0 1 0 0 ... 0 0 18 10 0 0 10 18 0 0
1 8 126 0 0 0 0 0 0 2 1 ... 0 56 0 12 0 27 15 0 16 0
3 6 93 0 1 0 2 0 0 0 0 ... 15 0 0 29 0 0 0 0 0 0

5 rows × 32 columns

To better understand the where clause features, let’s examine one of those features. The feature COUNT(sessions WHERE product_id_device = 5 and tablet), tells us how many sessions the customer purchased product_id 5 while on a tablet. Notice how the feature depends on multiple conditions (product_id = 5 & device = tablet).

[20]:
feature_matrix[["COUNT(sessions WHERE product_id_device = 5 and tablet)"]]
[20]:
COUNT(sessions WHERE product_id_device = 5 and tablet)
customer_id
2 1
5 0
4 1
1 0
3 0

Can I create an EntitySet using Dask or Spark dataframes? (BETA)#

Support for Dask EntitySets and Spark EntitySets is still in Beta - if you encounter any errors using either of these approaches, please let us know by creating a new issue on Github.

Yes! Featuretools supports creating an EntitySet from Dask dataframes or from Spark dataframes. You can simply follow the same process you would when creating an EntitySet from pandas dataframes.

There are some limitations to be aware of when using Dask or Spark dataframes. When creating a DataFrame, type inference can significantly slow down the runtime compared to pandas DataFrames, so users are encouraged to specify logical types for all columns during creation. Also, other quality checks are not performed, such as checking for unique index values. An EntitySet must be created entirely of one type of DataFrame (Dask, Spark, or pandas) - you cannot mix pandas DataFrames, Dask DataFrames, and Spark DataFrames with each other in the same EntitySet.

For more information on creating an EntitySet from Dask dataframes or from Spark dataframes, see the Using Dask EntitySets and the Using Spark EntitySets guides.

DFS#

Why is DFS not creating aggregation features?#

You may have created your EntitySet, and then applied DFS to create features. However, you may be puzzled as to why no aggregation features were created.

  • This is most likely because you have a single DataFrame in your EntitySet, and DFS is not capable of creating aggregation features with fewer than 2 DataFrames. Featuretools looks for a relationship, and aggregates based on that relationship.

Let’s look at a simple example.

[21]:
data = ft.demo.load_mock_customer()
transactions_df = data["transactions"].merge(data["sessions"]).merge(data["customers"])

es = ft.EntitySet(id="customer_data")
es = es.add_dataframe(
    dataframe_name="transactions", dataframe=transactions_df, index="transaction_id"
)
es
[21]:
Entityset: customer_data
  DataFrames:
    transactions [Rows: 500, Columns: 11]
  Relationships:
    No relationships

Notice how we only have 1 DataFrame in our EntitySet. If we try to create aggregation features on this EntitySet, it will not be possible because DFS needs 2 DataFrames to generate aggregation features.

[22]:
feature_matrix, feature_defs = ft.dfs(
    entityset=es, target_dataframe_name="transactions"
)
feature_defs
/home/docs/checkouts/readthedocs.org/user_builds/feature-labs-inc-featuretools/envs/v1.22.0/lib/python3.8/site-packages/featuretools/synthesis/deep_feature_synthesis.py:170: UserWarning: Only one dataframe in entityset, changing max_depth to 1 since deeper features cannot be created
  warnings.warn(
[22]:
[<Feature: session_id>,
 <Feature: product_id>,
 <Feature: amount>,
 <Feature: customer_id>,
 <Feature: device>,
 <Feature: zip_code>,
 <Feature: DAY(birthday)>,
 <Feature: DAY(join_date)>,
 <Feature: DAY(session_start)>,
 <Feature: DAY(transaction_time)>,
 <Feature: MONTH(birthday)>,
 <Feature: MONTH(join_date)>,
 <Feature: MONTH(session_start)>,
 <Feature: MONTH(transaction_time)>,
 <Feature: WEEKDAY(birthday)>,
 <Feature: WEEKDAY(join_date)>,
 <Feature: WEEKDAY(session_start)>,
 <Feature: WEEKDAY(transaction_time)>,
 <Feature: YEAR(birthday)>,
 <Feature: YEAR(join_date)>,
 <Feature: YEAR(session_start)>,
 <Feature: YEAR(transaction_time)>]

None of the above features are aggregation features. To fix this issue, you can add another DataFrame to your EntitySet.

Solution #1 - You can add new DataFrame if you have additional data.

[23]:
products_df = data["products"]
es = es.add_dataframe(
    dataframe_name="products", dataframe=products_df, index="product_id"
)
es
[23]:
Entityset: customer_data
  DataFrames:
    transactions [Rows: 500, Columns: 11]
    products [Rows: 5, Columns: 2]
  Relationships:
    No relationships

Notice how we now have an additional DataFrame in our EntitySet, called products.

Solution #2 - You can normalize an existing DataFrame.

[24]:
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", "join_date"],
    copy_columns=["session_start"],
)
es
[24]:
Entityset: customer_data
  DataFrames:
    transactions [Rows: 500, Columns: 7]
    products [Rows: 5, Columns: 2]
    sessions [Rows: 35, Columns: 6]
  Relationships:
    transactions.session_id -> sessions.session_id

Notice how we now have an additional DataFrame in our EntitySet, called sessions. Here, the normalization created a relationship between transactions and sessions. However, we could have specified a relationship between transactions and products if we had only used Solution #1.

Now, we can generate aggregation features.

[25]:
feature_matrix, feature_defs = ft.dfs(
    entityset=es, target_dataframe_name="transactions"
)
feature_defs[:-10]
[25]:
[<Feature: session_id>,
 <Feature: product_id>,
 <Feature: amount>,
 <Feature: DAY(birthday)>,
 <Feature: DAY(session_start)>,
 <Feature: DAY(transaction_time)>,
 <Feature: MONTH(birthday)>,
 <Feature: MONTH(session_start)>,
 <Feature: MONTH(transaction_time)>,
 <Feature: WEEKDAY(birthday)>,
 <Feature: WEEKDAY(session_start)>,
 <Feature: WEEKDAY(transaction_time)>,
 <Feature: YEAR(birthday)>,
 <Feature: YEAR(session_start)>,
 <Feature: YEAR(transaction_time)>,
 <Feature: sessions.device>,
 <Feature: sessions.customer_id>,
 <Feature: sessions.zip_code>,
 <Feature: sessions.COUNT(transactions)>,
 <Feature: sessions.MAX(transactions.amount)>,
 <Feature: sessions.MEAN(transactions.amount)>,
 <Feature: sessions.MIN(transactions.amount)>,
 <Feature: sessions.MODE(transactions.product_id)>,
 <Feature: sessions.NUM_UNIQUE(transactions.product_id)>,
 <Feature: sessions.SKEW(transactions.amount)>]

A few of the aggregation features are:

  • <Feature: sessions.MAX(transactions.amount)>

  • <Feature: sessions.SKEW(transactions.amount)>

  • <Feature: sessions.MIN(transactions.amount)>

  • <Feature: sessions.MEAN(transactions.amount)>

  • <Feature: sessions.COUNT(transactions)>

How do I speed up the runtime of DFS?#

One issue you may encounter while running ft.dfs is slow performance. While Featuretools has generally optimal default settings for calculating features, you may want to speed up performance when you are calculating on a large number of features.

One quick way to speed up performance is by adjusting the n_jobs settings of ft.dfs or ft.calculate_feature_matrix.

# setting n_jobs to -1 will use all cores

feature_matrix, feature_defs = ft.dfs(entityset=es,
                                      target_dataframe_name="customers",
                                      n_jobs=-1)


feature_matrix, feature_defs = ft.calculate_feature_matrix(entityset=es,
                                                           features=feature_defs,
                                                           n_jobs=-1)

For more ways to speed up performance, please visit:

How do I include only certain features when running DFS?#

When using DFS to generate features, you may wish to include only certain features. There are multiple ways that you do this:

  • Use ignore_columns to specify columns in a DataFrame that should not be used to create features. It is a dictionary mapping dataframe names to a list of column names to ignore.

  • Use drop_contains to drop features that contain any of the strings listed in this parameter.

  • Use drop_exact to drop features that exactly match any of the strings listed in this parameter.

Here is an example of using all three parameters:

[26]:
es = ft.demo.load_mock_customer(return_entityset=True)

feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    ignore_columns={
        "transactions": ["amount"],
        "customers": ["age", "gender", "birthday"],
    },  # ignore these columns
    drop_contains=["customers.SUM("],  # drop features that contain these strings
    drop_exact=["STD(transactions.quanity)"],
)  # drop features that exactly match

How do I specify primitives on a per column or per DataFrame basis?#

When using DFS to generate features, you may wish to use only certain features or DataFrames for specific primitives. This can be done through the primitive_options parameter. The primitive_options parameter is a dictionary that maps a primitive or a tuple of primitives to a dictionary containing options for the primitive(s). A primitive or tuple of primitives can also be mapped to a list of option dictionaries if the primitive(s) takes multiple inputs. The primitive keys can be the string names of the primitive, the primitive class, or specific instances of the primitive. Each dictionary supplies options for their respective input column. There are multiple ways to control how primitives get applied through these options:

  • Use ignore_dataframes to specify DataFrames that should not be used to create features for that primitive. It is a list of DataFrame names to ignore.

  • Use include_dataframes to specify the only DataFrames to be included to create features for that primitive. It is a list of DataFrame names to include.

  • Use ignore_columns to specify columns in a DataFrame that should not be used to create features for that primitive. It is a dictionary mapping a DataFrame name to a list of column names to ignore.

  • Use include_columns to specify the only columns in a DataFrame that should be used to create features for that primitive. It is a dictionary mapping a DataFrame name to a list of column names to include.

You can also use primitive_options to specify which DataFrames or columns you wish to use as groupbys for groupby transformation primitives:

  • Use ignore_groupby_dataframes to specify DataFrames that should not be used to get groupbys for that primitive. It is a list of DataFrame names to ignore.

  • Use include_groupby_dataframes to specify the only DataFrames that should be used to get groupbys for that primitive. It is a list of DataFrame names to include.

  • Use ignore_groupby_columns to specify columns in a DataFrame that should not be used as groupbys for that primitive. It is a dictionary mapping a DataFrame name to a list of column names to ignore.

  • Use include_groupby_columns to specify the only columns in a DataFrame that should be used as groupbys for that primitive. It is a dictionary mapping a DataFrame name to a list of column names to include.

Here is an example of using some of these options:

[27]:
es = ft.demo.load_mock_customer(return_entityset=True)

feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    primitive_options={
        "mode": {
            "ignore_dataframes": ["sessions"],
            "ignore_columns": {"products": ["brand"], "transactions": ["product_id"]},
        },
        # For mode, ignore the "sessions" DataFrame and only include "brands" in the
        # "products" dataframe and "product_id" in the "transactions" DataFrame
        ("count", "mean"): {"include_dataframes": ["sessions", "transactions"]}
        # For count and mean, only include the dataframes "sessions" and "transactions"
    },
)

Note that if options are given for a specific instance of a primitive and for the primitive generally (either by string name or class), the instances with their own options will not use the generic options. For example, in this case:

special_mean = Mean()
options = {
    special_mean: {'include_dataframes': ['customers']},
    'mean': {'include_dataframes': ['sessions']}

the primitive special_mean will not use the DataFrame sessions because it’s options have it only include customers. Every other instance of the Mean primitive will use the 'mean' options.

For more examples of specifying options for DFS, please visit:

If I didn’t specify the cutoff_time, what date will be used for the feature calculations?#

The cutoff time will be set to the current time using cutoff_time = datetime.now().

How do I select a certain amount of past data when calculating features?#

You may encounter a situation when you wish to make prediction using only a certain amount of historical data. You can accomplish this using the training_window parameter in ft.dfs. When you use the training_window, Featuretools will use the historical data between the cutoff_time and cutoff_time - training_window.

In order to make the calculation, Featuretools will check the time in the time_index column of the target_dataframe.

[28]:
es = ft.demo.load_mock_customer(return_entityset=True)
es["customers"].ww.time_index
[28]:
'join_date'

Our target_dataframe has a time_index, which is needed for the training_window calculation. Here, we are creating a cutoff time DataFrame so that we can have a unique training window for each customer.

[29]:
cutoff_times = pd.DataFrame()
cutoff_times["customer_id"] = [1, 2, 3, 1]
cutoff_times["time"] = pd.to_datetime(
    ["2014-1-1 04:00", "2014-1-1 05:00", "2014-1-1 06:00", "2014-1-1 08:00"]
)
cutoff_times["label"] = [True, True, False, True]

feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    cutoff_time=cutoff_times,
    cutoff_time_in_index=True,
    training_window="1 hour",
)
feature_matrix.head()
[29]:
zip_code COUNT(sessions) MODE(sessions.device) NUM_UNIQUE(sessions.device) COUNT(transactions) MAX(transactions.amount) MEAN(transactions.amount) MIN(transactions.amount) MODE(transactions.product_id) NUM_UNIQUE(transactions.product_id) ... STD(sessions.SUM(transactions.amount)) SUM(sessions.MAX(transactions.amount)) SUM(sessions.MEAN(transactions.amount)) SUM(sessions.MIN(transactions.amount)) SUM(sessions.NUM_UNIQUE(transactions.product_id)) SUM(sessions.SKEW(transactions.amount)) SUM(sessions.STD(transactions.amount)) MODE(transactions.sessions.device) NUM_UNIQUE(transactions.sessions.device) label
customer_id time
1 2014-01-01 04:00:00 60091 1 tablet 1 12 139.09 85.469167 6.78 4 5 ... NaN 139.09 85.469167 6.78 5.0 -0.830975 39.825249 tablet 1 True
2 2014-01-01 05:00:00 13244 1 tablet 1 13 118.85 77.304615 21.82 1 5 ... NaN 118.85 77.304615 21.82 5.0 -0.314918 33.725036 tablet 1 True
3 2014-01-01 06:00:00 13244 2 desktop 1 12 128.26 81.747500 20.06 3 5 ... 563.882303 220.02 172.597273 111.82 6.0 -0.289466 35.704680 desktop 1 False
1 2014-01-01 08:00:00 60091 1 mobile 1 16 126.11 88.755625 11.62 4 5 ... NaN 126.11 88.755625 11.62 5.0 -1.038434 32.324534 mobile 1 True

4 rows × 76 columns

Above, we ran DFS with training_window argument of 1 hour to create features that only used customer data collected in the last hour (from the cutoff time we provided).

Can I run DFS on a single table?#

Although possible, running DFS on a single table doesn’t make full use of DFS’s capabilities. For one, DFS will not be able to use any aggregation primitives, which require at least two tables. You will only be able to use transform primitives. This limits the complexity of the features that DFS can generate through feature stacking. Additionally, in certain situations, running single table DFS on data with time columns could risk label leakage. With data split in multiple tables, featuretools can filter data based on the cutoff time instead of assuming data was flattened appropriately, but it can not do this with only a single table.

If you only have a single table of data, DFS can certainly still be of use. There are two main ways to pass in a single table to DFS.

The first is to simply create an EntitySet with one table.

For example:

[30]:
transactions_df = ft.demo.load_mock_customer(return_single_table=True)

es = ft.EntitySet(id="customer_data")
es = es.add_dataframe(
    dataframe_name="transactions",
    dataframe=transactions_df,
    index="transaction_id",
    time_index="transaction_time",
)

feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="transactions",
    trans_primitives=[
        "time_since",
        "day",
        "is_weekend",
        "cum_min",
        "minute",
        "weekday",
        "percentile",
        "year",
        "week",
        "cum_mean",
    ],
)
/home/docs/checkouts/readthedocs.org/user_builds/feature-labs-inc-featuretools/envs/v1.22.0/lib/python3.8/site-packages/featuretools/synthesis/deep_feature_synthesis.py:170: UserWarning: Only one dataframe in entityset, changing max_depth to 1 since deeper features cannot be created
  warnings.warn(

The second way is to insert the dataframe into a dictionary mapping its name to a tuple containing specific dataframe information. We then pass in that dictionary to the dataframes argument in DFS.

In this scenario, for the value in our dictionary, we pass in a tuple containing the dataframe, its index column, and its time index. More information about the possible parameters can be found in the DFS documentation.

For example:

[31]:
transactions_df = ft.demo.load_mock_customer(return_single_table=True)

dataframes = {"transactions": (transactions_df, "transaction_id", "transaction_time")}

feature_matrix, feature_defs = ft.dfs(
    dataframes=dataframes,
    target_dataframe_name="transactions",
    trans_primitives=[
        "time_since",
        "day",
        "is_weekend",
        "cum_min",
        "minute",
        "weekday",
        "percentile",
        "year",
        "week",
        "cum_mean",
    ],
)
/home/docs/checkouts/readthedocs.org/user_builds/feature-labs-inc-featuretools/envs/v1.22.0/lib/python3.8/site-packages/featuretools/synthesis/deep_feature_synthesis.py:170: UserWarning: Only one dataframe in entityset, changing max_depth to 1 since deeper features cannot be created
  warnings.warn(

Before we examine the output, let’s look at our original single table.

[32]:
transactions_df.head()
[32]:
transaction_id session_id transaction_time product_id amount customer_id device session_start zip_code join_date birthday brand
298 298 1 2014-01-01 00:00:00 5 127.64 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18 A
2 2 1 2014-01-01 00:01:05 2 109.48 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18 B
308 308 1 2014-01-01 00:02:10 3 95.06 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18 B
116 116 1 2014-01-01 00:03:15 4 78.92 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18 B
371 371 1 2014-01-01 00:04:20 3 31.54 2 desktop 2014-01-01 13244 2012-04-15 23:31:04 1986-08-18 B

Now we can look at the transformations that Featuretools was able to apply to this single DataFrame to create feature matrix.

[33]:
feature_matrix.head()
[33]:
session_id product_id amount customer_id device zip_code brand CUM_MEAN(amount) CUM_MEAN(customer_id) CUM_MEAN(session_id) ... WEEK(session_start) WEEK(transaction_time) WEEKDAY(birthday) WEEKDAY(join_date) WEEKDAY(session_start) WEEKDAY(transaction_time) YEAR(birthday) YEAR(join_date) YEAR(session_start) YEAR(transaction_time)
transaction_id
298 1 5 127.64 2 desktop 13244 A 127.640000 2.0 1.0 ... 1 1 0 6 2 2 1986 2012 2014 2014
2 1 2 109.48 2 desktop 13244 B 118.560000 2.0 1.0 ... 1 1 0 6 2 2 1986 2012 2014 2014
308 1 3 95.06 2 desktop 13244 B 110.726667 2.0 1.0 ... 1 1 0 6 2 2 1986 2012 2014 2014
116 1 4 78.92 2 desktop 13244 B 102.775000 2.0 1.0 ... 1 1 0 6 2 2 1986 2012 2014 2014
371 1 3 31.54 2 desktop 13244 B 88.528000 2.0 1.0 ... 1 1 0 6 2 2 1986 2012 2014 2014

5 rows × 44 columns

How do I prevent label leakage with DFS?#

One concern you might have with using DFS is about label leakage. You want to make sure that labels in your data aren’t used incorrectly to create features and the feature matrix.

Featuretools is particularly focused on helping users avoid label leakage.

There are two ways to prevent label leakage depending on if your data has timestamps or not.

1. Data without timestamps#

In the case where you do not have timestamps, you can create one EntitySet using only the training data and then run ft.dfs. This will create a feature matrix using only the training data, but also return a list of feature definitions. Next, you can create an EntitySet using the test data and recalculate the same features by calling ft.calculate_feature_matrix with the list of feature definitions from before.

Here is what that flow would look like:

First, let’s create our training data.

[34]:
train_data = pd.DataFrame(
    {
        "customer_id": [1, 2, 3, 4, 5],
        "age": [40, 50, 10, 20, 30],
        "gender": ["m", "f", "m", "f", "f"],
        "signup_date": pd.date_range("2014-01-01 01:41:50", periods=5, freq="25min"),
        "labels": [True, False, True, False, True],
    }
)
train_data.head()
[34]:
customer_id age gender signup_date labels
0 1 40 m 2014-01-01 01:41:50 True
1 2 50 f 2014-01-01 02:06:50 False
2 3 10 m 2014-01-01 02:31:50 True
3 4 20 f 2014-01-01 02:56:50 False
4 5 30 f 2014-01-01 03:21:50 True

Now, we can create an entityset for our training data.

[35]:
es_train_data = ft.EntitySet(id="customer_train_data")
es_train_data = es_train_data.add_dataframe(
    dataframe_name="customers", dataframe=train_data, index="customer_id"
)
es_train_data
[35]:
Entityset: customer_train_data
  DataFrames:
    customers [Rows: 5, Columns: 5]
  Relationships:
    No relationships

Next, we are ready to create our features, and feature matrix for the training data. We don’t want Featuretools to use the labels column to build new features, so we will use the ignore_columns option to exclude it. This would also remove the labels column from the feature matrix, so we will tell DFS to include it as a seed feature.

[36]:
labels_feature = ft.Feature(es_train_data["customers"].ww["labels"])
feature_matrix_train, feature_defs = ft.dfs(
    entityset=es_train_data,
    target_dataframe_name="customers",
    ignore_columns={"customers": ["labels"]},
    seed_features=[labels_feature],
)
feature_matrix_train
/home/docs/checkouts/readthedocs.org/user_builds/feature-labs-inc-featuretools/envs/v1.22.0/lib/python3.8/site-packages/featuretools/synthesis/deep_feature_synthesis.py:170: UserWarning: Only one dataframe in entityset, changing max_depth to 1 since deeper features cannot be created
  warnings.warn(
[36]:
age labels DAY(signup_date) MONTH(signup_date) WEEKDAY(signup_date) YEAR(signup_date)
customer_id
1 40 True 1 1 2 2014
2 50 False 1 1 2 2014
3 10 True 1 1 2 2014
4 20 False 1 1 2 2014
5 30 True 1 1 2 2014

We will also encode our feature matrix to make machine learning compatible features.

[37]:
feature_matrix_train_enc, features_enc = ft.encode_features(
    feature_matrix_train, feature_defs
)
feature_matrix_train_enc.head()
[37]:
age labels DAY(signup_date) = 1 DAY(signup_date) is unknown MONTH(signup_date) = 1 MONTH(signup_date) is unknown WEEKDAY(signup_date) = 2 WEEKDAY(signup_date) is unknown YEAR(signup_date) = 2014 YEAR(signup_date) is unknown
customer_id
1 40 True True False True False True False True False
2 50 False True False True False True False True False
3 10 True True False True False True False True False
4 20 False True False True False True False True False
5 30 True True False True False True False True False

Notice how the whole feature matrix only includes numeric and boolean values now.

Now we can use the feature definitions to calculate our feature matrix for the test data, and avoid label leakage.

[38]:
test_train = pd.DataFrame(
    {
        "customer_id": [6, 7, 8, 9, 10],
        "age": [20, 25, 55, 22, 35],
        "gender": ["f", "m", "m", "m", "m"],
        "signup_date": pd.date_range("2014-01-01 01:41:50", periods=5, freq="25min"),
        "labels": [True, False, False, True, True],
    }
)

es_test_data = ft.EntitySet(id="customer_test_data")
es_test_data = es_test_data.add_dataframe(
    dataframe_name="customers",
    dataframe=test_train,
    index="customer_id",
    time_index="signup_date",
)

# Use the feature definitions from earlier
feature_matrix_enc_test = ft.calculate_feature_matrix(
    features=features_enc, entityset=es_test_data
)

feature_matrix_enc_test.head()
[38]:
age labels DAY(signup_date) = 1 DAY(signup_date) is unknown MONTH(signup_date) = 1 MONTH(signup_date) is unknown WEEKDAY(signup_date) = 2 WEEKDAY(signup_date) is unknown YEAR(signup_date) = 2014 YEAR(signup_date) is unknown
customer_id
6 20 True True False True False True False True False
7 25 False True False True False True False True False
8 55 False True False True False True False True False
9 22 True True False True False True False True False
10 35 True True False True False True False True False

Check out the Modeling section for an example of using the encoded matrix with sklearn.

2. Data with timestamps#

If your data has timestamps, the best way to prevent label leakage is to use a list of cutoff times, which specify the last point in time data is allowed to be used for each row in the resulting feature matrix. To use cutoff times, you need to set a time index for each time sensitive DataFrame in your entity set.

Tip: Even if your data doesn’t have time stamps, you could add a column with dummy timestamps that can be used by Featuretools as time index.

When you call ft.dfs, you can provide a DataFrame of cutoff times like this:

[39]:
cutoff_times = pd.DataFrame(
    {
        "customer_id": [1, 2, 3, 4, 5],
        "time": pd.date_range("2014-01-01 01:41:50", periods=5, freq="25min"),
    }
)
cutoff_times.head()
[39]:
customer_id time
0 1 2014-01-01 01:41:50
1 2 2014-01-01 02:06:50
2 3 2014-01-01 02:31:50
3 4 2014-01-01 02:56:50
4 5 2014-01-01 03:21:50
[40]:
train_test_data = pd.DataFrame(
    {
        "customer_id": [1, 2, 3, 4, 5],
        "age": [20, 25, 55, 22, 35],
        "gender": ["f", "m", "m", "m", "m"],
        "signup_date": pd.date_range("2010-01-01 01:41:50", periods=5, freq="25min"),
    }
)

es_train_test_data = ft.EntitySet(id="customer_train_test_data")
es_train_test_data = es_train_test_data.add_dataframe(
    dataframe_name="customers",
    dataframe=train_test_data,
    index="customer_id",
    time_index="signup_date",
)

feature_matrix_train_test, features = ft.dfs(
    entityset=es_train_test_data,
    target_dataframe_name="customers",
    cutoff_time=cutoff_times,
    cutoff_time_in_index=True,
)
feature_matrix_train_test.head()
/home/docs/checkouts/readthedocs.org/user_builds/feature-labs-inc-featuretools/envs/v1.22.0/lib/python3.8/site-packages/featuretools/synthesis/deep_feature_synthesis.py:170: UserWarning: Only one dataframe in entityset, changing max_depth to 1 since deeper features cannot be created
  warnings.warn(
[40]:
age DAY(signup_date) MONTH(signup_date) WEEKDAY(signup_date) YEAR(signup_date)
customer_id time
1 2014-01-01 01:41:50 20 1 1 4 2010
2 2014-01-01 02:06:50 25 1 1 4 2010
3 2014-01-01 02:31:50 55 1 1 4 2010
4 2014-01-01 02:56:50 22 1 1 4 2010
5 2014-01-01 03:21:50 35 1 1 4 2010

Above, we have created a feature matrix that uses cutoff times to avoid label leakage. We could also encode this feature matrix using ft.encode_features.

What is the difference between passing a primitive object versus a string to DFS?#

There are 2 ways to pass primitives to DFS: the primitive object, or a string of the primitive name.

We will use the Transform primitive called TimeSincePrevious to illustrate the differences.

First, let’s use the string of primitive name.

[41]:
es = ft.demo.load_mock_customer(return_entityset=True)
[42]:
feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    agg_primitives=[],
    trans_primitives=["time_since_previous"],
)
feature_matrix
[42]:
zip_code TIME_SINCE_PREVIOUS(join_date)
customer_id
5 60091 NaN
4 60091 22948824.0
1 60091 744019.0
3 13244 10212841.0
2 13244 21282510.0

Now, let’s use the primitive object.

[43]:
from featuretools.primitives import TimeSincePrevious

feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    agg_primitives=[],
    trans_primitives=[TimeSincePrevious],
)
feature_matrix
[43]:
zip_code TIME_SINCE_PREVIOUS(join_date)
customer_id
5 60091 NaN
4 60091 22948824.0
1 60091 744019.0
3 13244 10212841.0
2 13244 21282510.0

As we can see above, the feature matrix is the same.

However, if we need to modify controllable parameters in the primitive, we should use the primitive object. For instance, let’s make TimeSincePrevious return units of hours (the default is in seconds).

[44]:
from featuretools.primitives import TimeSincePrevious

time_since_previous_in_hours = TimeSincePrevious(unit="hours")

feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    agg_primitives=[],
    trans_primitives=[time_since_previous_in_hours],
)
feature_matrix
[44]:
zip_code TIME_SINCE_PREVIOUS(join_date, unit=hours)
customer_id
5 60091 NaN
4 60091 6374.673333
1 60091 206.671944
3 13244 2836.900278
2 13244 5911.808333

Features#

How can I select features based on some attributes (a specific string, an explicit primitive type, a return type, a given depth)?#

You may wish to select a subset of your features based on some attributes.

Let’s say you wanted to select features that had the string amount in its name. You can check for this by using the get_name function on the feature definitions.

[45]:
es = ft.demo.load_mock_customer(return_entityset=True)

feature_defs = ft.dfs(
    entityset=es, target_dataframe_name="customers", features_only=True
)

features_with_amount = []
for x in feature_defs:
    if "amount" in x.get_name():
        features_with_amount.append(x)
features_with_amount[0:5]
[45]:
[<Feature: MAX(transactions.amount)>,
 <Feature: MEAN(transactions.amount)>,
 <Feature: MIN(transactions.amount)>,
 <Feature: SKEW(transactions.amount)>,
 <Feature: STD(transactions.amount)>]

You might also want to only select features that are aggregation features.

[46]:
from featuretools import AggregationFeature

features_only_aggregations = []
for x in feature_defs:
    if type(x) == AggregationFeature:
        features_only_aggregations.append(x)
features_only_aggregations[0:5]
[46]:
[<Feature: COUNT(sessions)>,
 <Feature: MODE(sessions.device)>,
 <Feature: NUM_UNIQUE(sessions.device)>,
 <Feature: COUNT(transactions)>,
 <Feature: MAX(transactions.amount)>]

Also, you might only want to select features that are calculated at a certain depth. You can do this by using the get_depth function.

[47]:
features_only_depth_2 = []
for x in feature_defs:
    if x.get_depth() == 2:
        features_only_depth_2.append(x)
features_only_depth_2[0:5]
[47]:
[<Feature: MAX(sessions.COUNT(transactions))>,
 <Feature: MAX(sessions.MEAN(transactions.amount))>,
 <Feature: MAX(sessions.MIN(transactions.amount))>,
 <Feature: MAX(sessions.NUM_UNIQUE(transactions.product_id))>,
 <Feature: MAX(sessions.SKEW(transactions.amount))>]

Finally, you might only want features that return a certain type. You can do this by using the column_schema attribute. For more information on working with column schemas, take a look at Transitioning from Variables to Woodwork.

[48]:
features_only_numeric = []
for x in feature_defs:
    if "numeric" in x.column_schema.semantic_tags:
        features_only_numeric.append(x)
features_only_numeric[0:5]
[48]:
[<Feature: COUNT(sessions)>,
 <Feature: NUM_UNIQUE(sessions.device)>,
 <Feature: COUNT(transactions)>,
 <Feature: MAX(transactions.amount)>,
 <Feature: MEAN(transactions.amount)>]

Once you have your specific feature list, you can use ft.calculate_feature_matrix to generate a feature matrix for only those features.

For our example, let’s use the features with only the string amount in its name.

[49]:
feature_matrix = ft.calculate_feature_matrix(
    entityset=es, features=features_with_amount
)  # change to your specific feature list
feature_matrix.head()
[49]:
MAX(transactions.amount) MEAN(transactions.amount) MIN(transactions.amount) SKEW(transactions.amount) STD(transactions.amount) SUM(transactions.amount) MAX(sessions.MEAN(transactions.amount)) MAX(sessions.MIN(transactions.amount)) MAX(sessions.SKEW(transactions.amount)) MAX(sessions.STD(transactions.amount)) ... STD(sessions.MAX(transactions.amount)) STD(sessions.MEAN(transactions.amount)) STD(sessions.MIN(transactions.amount)) STD(sessions.SKEW(transactions.amount)) STD(sessions.SUM(transactions.amount)) SUM(sessions.MAX(transactions.amount)) SUM(sessions.MEAN(transactions.amount)) SUM(sessions.MIN(transactions.amount)) SUM(sessions.SKEW(transactions.amount)) SUM(sessions.STD(transactions.amount))
customer_id
5 149.02 80.375443 7.55 -0.025941 44.095630 6349.66 94.481667 20.65 0.602209 51.149250 ... 7.928001 11.007471 4.961414 0.415426 402.775486 839.76 472.231119 86.49 0.014384 259.873954
4 149.95 80.070459 5.73 -0.036348 45.068765 8727.68 110.450000 54.83 0.382868 54.293903 ... 3.514421 13.027258 16.960575 0.387884 235.992478 1157.99 649.657515 131.51 0.002764 356.125829
1 139.43 71.631905 5.81 0.019698 40.442059 9025.62 88.755625 26.36 0.640252 46.905665 ... 7.322191 13.759314 6.954507 0.589386 279.510713 1057.97 582.193117 78.59 -0.476122 312.745952
3 149.15 67.060430 5.89 0.418230 43.683296 6236.62 82.109444 20.06 0.854976 50.110120 ... 10.724241 11.174282 5.424407 0.429374 219.021420 847.63 405.237462 66.21 2.286086 257.299895
2 146.81 77.422366 8.73 0.098259 37.705178 7200.28 96.581000 56.46 0.755711 47.935920 ... 17.221593 11.477071 15.874374 0.509798 251.609234 931.63 548.905851 154.60 -0.277640 258.700528

5 rows × 37 columns

Above, notice how all the column names for our feature matrix contain the string amount.

How do I create where features?#

Sometimes, you might want to create features that are conditioned on a second value before it is calculated. This extra filter is called a “where clause”. You can create these features using the using the interesting_values of a column.

If you have categorical columns in your EntitySet, you can use add_interesting_values. This function will find interesting values for your categorical columns, which can then be used to generate “where” clauses.

First, let’s create our EntitySet.

[50]:
es = ft.demo.load_mock_customer(return_entityset=True)
es
[50]:
Entityset: transactions
  DataFrames:
    transactions [Rows: 500, Columns: 6]
    products [Rows: 5, Columns: 3]
    sessions [Rows: 35, Columns: 5]
    customers [Rows: 5, Columns: 5]
  Relationships:
    transactions.product_id -> products.product_id
    transactions.session_id -> sessions.session_id
    sessions.customer_id -> customers.customer_id

Now we can add the interesting values for the categorical column.

[51]:
es.add_interesting_values()

Now we can run DFS with the where_primitives argument to define which primitives to apply with where clauses. In this case, let’s use the primitive count. For this to work, the primitive count must be present in both agg_primitives and where_primitives.

[52]:
feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    agg_primitives=["count"],
    where_primitives=["count"],
    trans_primitives=[],
)
feature_matrix.head()
[52]:
zip_code COUNT(sessions) COUNT(transactions) COUNT(sessions WHERE device = tablet) COUNT(sessions WHERE device = desktop) COUNT(sessions WHERE device = mobile) COUNT(sessions WHERE customers.zip_code = 13244) COUNT(sessions WHERE customers.zip_code = 60091) COUNT(transactions WHERE sessions.device = tablet) COUNT(transactions WHERE sessions.device = desktop) COUNT(transactions WHERE sessions.device = mobile)
customer_id
5 60091 6 79 1 2 3 0 6 14 29 36
4 60091 8 109 1 3 4 0 8 18 38 53
1 60091 8 126 3 2 3 0 8 43 27 56
3 13244 6 93 1 4 1 6 0 15 62 16
2 13244 7 93 2 3 2 7 0 28 34 31

We have now created some useful features. One example of a useful feature is the COUNT(sessions WHERE device = tablet). This feature tells us how many sessions a customer completed on a tablet.

[53]:
feature_matrix[["COUNT(sessions WHERE device = tablet)"]]
[53]:
COUNT(sessions WHERE device = tablet)
customer_id
5 1
4 1
1 3
3 1
2 2

Primitives#

What is the difference between the primitive types (Transform, GroupBy Transform, & Aggregation)?#

You might curious to know the difference between the primitive groups. Let’s review the differences between transform, groupby transform, and aggregation primitives.

First, let’s create a simple EntitySet.

[54]:
import pandas as pd
import featuretools as ft

df = pd.DataFrame(
    {
        "id": [1, 2, 3, 4, 5, 6],
        "time_index": pd.date_range("1/1/2019", periods=6, freq="D"),
        "group": ["a", "a", "a", "a", "a", "a"],
        "val": [5, 1, 10, 20, 6, 23],
    }
)
es = ft.EntitySet()
es = es.add_dataframe(
    dataframe_name="observations", dataframe=df, index="id", time_index="time_index"
)

es = es.normalize_dataframe(
    base_dataframe_name="observations", new_dataframe_name="groups", index="group"
)

es.plot()
[54]:
../_images/resources_frequently_asked_questions_119_0.svg

After calling normalize_dataframe, the column “group” has the semantic tag “foreign_key” because it identifies another DataFrame. Alternatively, it could be set using the semantic_tags parameter when we first call es.add_dataframe().

Transform Primitive#

The cum_sum primitive calculates the running sum in list of numbers.

[55]:
from featuretools.primitives import CumSum

cum_sum = CumSum()
cum_sum([1, 2, 3, 4, 5]).tolist()
[55]:
[1, 3, 6, 10, 15]

If we apply it using the trans_primitives argument it will calculate it over the entire observations DataFrame like this:

[56]:
feature_matrix, feature_defs = ft.dfs(
    target_dataframe_name="observations",
    entityset=es,
    agg_primitives=[],
    trans_primitives=["cum_sum"],
    groupby_trans_primitives=[],
)

feature_matrix
[56]:
group val CUM_SUM(val)
id
1 a 5 5.0
2 a 1 6.0
3 a 10 16.0
4 a 20 36.0
5 a 6 42.0
6 a 23 65.0

Groupby Transform Primitive#

If we apply it using groupby_trans_primitives, then DFS will first group by any foreign key columns before applying the transform primitive. As a result, we get the cumulative sum by group.

[57]:
feature_matrix, feature_defs = ft.dfs(
    target_dataframe_name="observations",
    entityset=es,
    agg_primitives=[],
    trans_primitives=[],
    groupby_trans_primitives=["cum_sum"],
)

feature_matrix
[57]:
group val CUM_SUM(val) by group
id
1 a 5 5.0
2 a 1 6.0
3 a 10 16.0
4 a 20 36.0
5 a 6 42.0
6 a 23 65.0

Aggregation Primitive#

Finally, there is also the aggregation primitive “sum”. If we use sum, it will calculate the sum for the group at the cutoff time for each row. Because we didn’t specify a cutoff time it will use all the data for each group for each row.

[58]:
feature_matrix, feature_defs = ft.dfs(
    target_dataframe_name="observations",
    entityset=es,
    agg_primitives=["sum"],
    trans_primitives=[],
    cutoff_time_in_index=True,
    groupby_trans_primitives=[],
)

feature_matrix
[58]:
group val groups.SUM(observations.val)
id time
1 2023-01-31 22:29:40.693015 a 5 65.0
2 2023-01-31 22:29:40.693015 a 1 65.0
3 2023-01-31 22:29:40.693015 a 10 65.0
4 2023-01-31 22:29:40.693015 a 20 65.0
5 2023-01-31 22:29:40.693015 a 6 65.0
6 2023-01-31 22:29:40.693015 a 23 65.0

If we set the cutoff time of each row to be the time index, then use sum as an aggregation primitive, the result is the same as cum_sum. (Though the order is different in the displayed dataframe).

[59]:
cutoff_time = df[["id", "time_index"]]
cutoff_time
[59]:
id time_index
1 1 2019-01-01
2 2 2019-01-02
3 3 2019-01-03
4 4 2019-01-04
5 5 2019-01-05
6 6 2019-01-06
[60]:
feature_matrix, feature_defs = ft.dfs(
    target_dataframe_name="observations",
    entityset=es,
    agg_primitives=["sum"],
    trans_primitives=[],
    groupby_trans_primitives=[],
    cutoff_time_in_index=True,
    cutoff_time=cutoff_time,
)

feature_matrix
[60]:
group val groups.SUM(observations.val)
id time
1 2019-01-01 a 5 5.0
2 2019-01-02 a 1 6.0
3 2019-01-03 a 10 16.0
4 2019-01-04 a 20 36.0
5 2019-01-05 a 6 42.0
6 2019-01-06 a 23 65.0

How do I get a list of all Aggregation and Transform primitives?#

You can do featuretools.list_primitives() to get all the primitive in Featuretools. It will return a DataFrame with the names, type, and description of the primitives, and if the primitive can be used with entitysets created from Dask dataframes.

[61]:
df_primitives = ft.list_primitives()
df_primitives.head()
[61]:
name type dask_compatible spark_compatible description valid_inputs return_type
0 time_since_last_min aggregation False False Calculates the time since the minimum value oc... <ColumnSchema (Logical Type = Datetime) (Seman... <ColumnSchema (Logical Type = Double) (Semanti...
1 count_less_than aggregation False False Determines the number of values less than a co... <ColumnSchema (Semantic Tags = ['numeric'])> <ColumnSchema (Logical Type = Integer) (Semant...
2 all aggregation True False Calculates if all values are 'True' in a list. <ColumnSchema (Logical Type = BooleanNullable)... <ColumnSchema (Logical Type = Boolean)>
3 percent_true aggregation True False Determines the percent of `True` values. <ColumnSchema (Logical Type = BooleanNullable)... <ColumnSchema (Logical Type = Double) (Semanti...
4 any aggregation True False Determines if any value is 'True' in a list. <ColumnSchema (Logical Type = BooleanNullable)... <ColumnSchema (Logical Type = Boolean)>
[62]:
df_primitives.tail()
[62]:
name type dask_compatible spark_compatible description valid_inputs return_type
172 greater_than_equal_to_scalar transform True True Determines if values are greater than or equal... <ColumnSchema (Semantic Tags = ['numeric'])> <ColumnSchema (Logical Type = BooleanNullable)>
173 is_year_end transform True True Determines if a date falls on the end of a year. <ColumnSchema (Logical Type = Datetime)> <ColumnSchema (Logical Type = BooleanNullable)>
174 stopword_count transform False False Determines number of stopwords in a string. <ColumnSchema (Logical Type = NaturalLanguage)> <ColumnSchema (Logical Type = IntegerNullable)...
175 isin transform True True Determines whether a value is present in a pro... <ColumnSchema> <ColumnSchema (Logical Type = Boolean)>
176 multiply_boolean transform True False Performs element-wise multiplication of two li... <ColumnSchema (Logical Type = BooleanNullable)... <ColumnSchema (Logical Type = BooleanNullable)>

What primitives can I use when creating a feature matrix from a Dask EntitySet? (BETA)#

Support for Dask EntitySets is still in Beta - if you encounter any errors using this approach, please let us know by creating a new issue on Github.

When creating a feature matrix from a Dask EntitySet, only certain primitives can be used. Computation of certain features is quite expensive in a distributed environment, and as a result only a subset of Featuretools primitives are currently supported when using a Dask EntitySet.

The table returned by featuretools.list_primitives() will contain a column labeled dask_compatible. Any primitive that has a value of True in this column can be used safely when computing a feature matrix from a Dask EntitySet.

How do I change the units for a TimeSince primitive?#

There are a few primitives in Featuretools that make some time-based calculation. These include TimeSince, TimeSincePrevious, TimeSinceLast, TimeSinceFirst.

You can change the units from the default seconds to any valid time unit, by doing the following:

[63]:
from featuretools.primitives import (
    TimeSince,
    TimeSincePrevious,
    TimeSinceLast,
    TimeSinceFirst,
)

time_since = TimeSince(unit="minutes")
time_since_previous = TimeSincePrevious(unit="hours")
time_since_last = TimeSinceLast(unit="days")
time_since_first = TimeSinceFirst(unit="years")

es = ft.demo.load_mock_customer(return_entityset=True)

feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    agg_primitives=[time_since_last, time_since_first],
    trans_primitives=[time_since, time_since_previous],
)

Above, we changed the units to the following: - minutes for TimeSince - hours for TimeSincePrevious - days for TimeSinceLast - years for TimeSinceFirst.

Now we can see that our feature matrix contains multiple features where the units for the TimeSince primitives are changed.

[64]:
feature_matrix.head()
[64]:
zip_code TIME_SINCE_FIRST(sessions.session_start, unit=years) TIME_SINCE_LAST(sessions.session_start, unit=days) TIME_SINCE_FIRST(transactions.transaction_time, unit=years) TIME_SINCE_LAST(transactions.transaction_time, unit=days) TIME_SINCE(birthday, unit=minutes) TIME_SINCE(join_date, unit=minutes) TIME_SINCE_PREVIOUS(join_date, unit=hours) TIME_SINCE_FIRST(transactions.sessions.session_start, unit=years) TIME_SINCE_LAST(transactions.sessions.session_start, unit=days)
customer_id
5 60091 9.089053 3317.602500 9.089053 3317.597234 2.025639e+07 6.597662e+06 NaN 9.089053 3317.602500
4 60091 9.089033 3317.713843 9.089033 3317.707072 8.660070e+06 6.215181e+06 6374.673333 9.089033 3317.713843
1 60091 9.089002 3317.638612 9.089002 3317.627327 1.501191e+07 6.202781e+06 206.671944 9.089002 3317.638612
3 13244 9.088897 3317.573160 9.088897 3317.561875 1.009719e+07 6.032567e+06 2836.900278 9.088897 3317.573160
2 13244 9.089086 3317.596482 9.089086 3317.587454 1.917495e+07 5.677859e+06 5911.808333 9.089086 3317.596482

There are now features where time unit is different from the default of seconds, such as TIME_SINCE_LAST(sessions.session_start, unit=days), and TIME_SINCE_FIRST(sessions.session_start, unit=years).

Modeling#

How does my train & test data work with Featuretools and sklearn’s train_test_split?#

You might be wondering how to properly use your train & test data with Featuretools, and sklearn’s train_test_split. There are a few things you must do to ensure accuracy with this workflow.

Let’s imagine we have a Dataframes for our train data, with the labels.

[65]:
train_data = pd.DataFrame(
    {
        "customer_id": [1, 2, 3, 4, 5],
        "age": [20, 25, 55, 22, 35],
        "gender": ["f", "m", "m", "m", "m"],
        "signup_date": pd.date_range("2010-01-01 01:41:50", periods=5, freq="25min"),
        "labels": [False, True, True, False, False],
    }
)
train_data.head()
[65]:
customer_id age gender signup_date labels
0 1 20 f 2010-01-01 01:41:50 False
1 2 25 m 2010-01-01 02:06:50 True
2 3 55 m 2010-01-01 02:31:50 True
3 4 22 m 2010-01-01 02:56:50 False
4 5 35 m 2010-01-01 03:21:50 False

Now we can create our EntitySet for the train data, and create our features. To prevent label leakage, we will use cutoff times (see earlier question).

[66]:
es_train_data = ft.EntitySet(id="customer_data")
es_train_data = es_train_data.add_dataframe(
    dataframe_name="customers", dataframe=train_data, index="customer_id"
)

cutoff_times = pd.DataFrame(
    {
        "customer_id": [1, 2, 3, 4, 5],
        "time": pd.date_range("2014-01-01 01:41:50", periods=5, freq="25min"),
    }
)

feature_matrix_train, features = ft.dfs(
    entityset=es_train_data,
    target_dataframe_name="customers",
    cutoff_time=cutoff_times,
    cutoff_time_in_index=True,
)
feature_matrix_train.head()
/home/docs/checkouts/readthedocs.org/user_builds/feature-labs-inc-featuretools/envs/v1.22.0/lib/python3.8/site-packages/featuretools/synthesis/deep_feature_synthesis.py:170: UserWarning: Only one dataframe in entityset, changing max_depth to 1 since deeper features cannot be created
  warnings.warn(
[66]:
age labels DAY(signup_date) MONTH(signup_date) WEEKDAY(signup_date) YEAR(signup_date)
customer_id time
1 2014-01-01 01:41:50 20 False 1 1 4 2010
2 2014-01-01 02:06:50 25 True 1 1 4 2010
3 2014-01-01 02:31:50 55 True 1 1 4 2010
4 2014-01-01 02:56:50 22 False 1 1 4 2010
5 2014-01-01 03:21:50 35 False 1 1 4 2010

We will also encode our feature matrix to compatible for machine learning algorithms.

[67]:
feature_matrix_train_enc, feature_enc = ft.encode_features(
    feature_matrix_train, features
)
feature_matrix_train_enc.head()
[67]:
age labels DAY(signup_date) = 1 DAY(signup_date) is unknown MONTH(signup_date) = 1 MONTH(signup_date) is unknown WEEKDAY(signup_date) = 4 WEEKDAY(signup_date) is unknown YEAR(signup_date) = 2010 YEAR(signup_date) is unknown
customer_id time
1 2014-01-01 01:41:50 20 False True False True False True False True False
2 2014-01-01 02:06:50 25 True True False True False True False True False
3 2014-01-01 02:31:50 55 True True False True False True False True False
4 2014-01-01 02:56:50 22 False True False True False True False True False
5 2014-01-01 03:21:50 35 False True False True False True False True False
[68]:
from sklearn.model_selection import train_test_split

X = feature_matrix_train_enc.drop(["labels"], axis=1)
y = feature_matrix_train_enc["labels"]

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.25)

Now you can use the encoded feature matrix with sklearn’s train_test_split. This will allow you to train your model, and tune your parameters.

How are categorical columns encoded when splitting training and testing data?#

You might be wondering what happens when categorical columns are encoded with your training and testing data. You might be curious to know what happens if the train data has a categorical column that is not present in the testing data.

Let’s explore a simple example to see what happens during the encoding process.

[69]:
train_data = pd.DataFrame(
    {
        "customer_id": [1, 2, 3, 4, 5],
        "product_purchased": ["coke zero", "car", "toothpaste", "coke zero", "car"],
    }
)
es_train = ft.EntitySet(id="customer_data")
es_train = es_train.add_dataframe(
    dataframe_name="customers",
    dataframe=train_data,
    index="customer_id",
    logical_types={"product_purchased": ww.logical_types.Categorical},
)
feature_matrix_train, features = ft.dfs(
    entityset=es_train, target_dataframe_name="customers"
)
feature_matrix_train
/home/docs/checkouts/readthedocs.org/user_builds/feature-labs-inc-featuretools/envs/v1.22.0/lib/python3.8/site-packages/featuretools/synthesis/deep_feature_synthesis.py:170: UserWarning: Only one dataframe in entityset, changing max_depth to 1 since deeper features cannot be created
  warnings.warn(
[69]:
product_purchased
customer_id
1 coke zero
2 car
3 toothpaste
4 coke zero
5 car

We will use ft.encode_features to properly encode the product_purchased column.

[70]:
feature_matrix_train_encoded, features_encoded = ft.encode_features(
    feature_matrix_train, features
)
feature_matrix_train_encoded.head()
[70]:
product_purchased = coke zero product_purchased = car product_purchased = toothpaste product_purchased is unknown
customer_id
1 True False False False
2 False True False False
3 False False True False
4 True False False False
5 False True False False

Now lets imagine we have some test data that has doesn’t have one of the categorical values (toothpaste). Also, the test data has a value that wasn’t present in the train data (water).

[71]:
test_data = pd.DataFrame(
    {
        "customer_id": [6, 7, 8, 9, 10],
        "product_purchased": ["coke zero", "car", "coke zero", "coke zero", "water"],
    }
)

es_test = ft.EntitySet(id="customer_data")
es_test = es_test.add_dataframe(
    dataframe_name="customers", dataframe=test_data, index="customer_id"
)

feature_matrix_test = ft.calculate_feature_matrix(
    entityset=es_test, features=features_encoded
)
feature_matrix_test.head()
[71]:
product_purchased = coke zero product_purchased = car product_purchased = toothpaste product_purchased is unknown
customer_id
6 True False False False
7 False True False False
8 True False False False
9 True False False False
10 False False False True

As seen above, we were able to successfully handle the encoding, and deal with the following complications: - toothpaste was present in the training data but not present in the testing data - water was present in the test data but not present in the training data.

Errors & Warnings#

Why am I getting this error ‘Index is not unique on dataframe’?#

You may be trying to create your EntitySet, and run into this error.

IndexError: Index column must be unique

This is because each dataframe in your EntitySet needs a unique index.

Let’s look at a simple example.

[72]:
product_df = pd.DataFrame({"id": [1, 2, 3, 4, 4], "rating": [3.5, 4.0, 4.5, 1.5, 5.0]})
product_df
[72]:
id rating
0 1 3.5
1 2 4.0
2 3 4.5
3 4 1.5
4 4 5.0

Notice how the id column has a duplicate index of 4. If you try to add this dataframe to the EntitySet, you will run into the following error.

es = ft.EntitySet(id="product_data")
es = es.add_dataframe(dataframe_name="products",
                      dataframe=product_df,
                      index="id")
---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
<ipython-input-78-854fbaf207f8> in <module>
      1 es = ft.EntitySet(id="product_data")
----> 2 es = es.add_dataframe(dataframe_name="products",
      3                       dataframe=product_df,
      4                       index="id")

~/Code/featuretools/featuretools/entityset/entityset.py in add_dataframe(self, dataframe, dataframe_name, index, logical_types, semantic_tags, make_index, time_index, secondary_time_index, already_sorted)
    625             index_was_created, index, dataframe = _get_or_create_index(index, make_index, dataframe)
    626
--> 627             dataframe.ww.init(name=dataframe_name,
    628                               index=index,
    629                               time_index=time_index,

/usr/local/Caskroom/miniconda/base/envs/featuretools/lib/python3.8/site-packages/woodwork/table_accessor.py in init(self, index, time_index, logical_types, already_sorted, schema, validate, use_standard_tags, **kwargs)
     94         """
     95         if validate:
---> 96             _validate_accessor_params(self._dataframe, index, time_index, logical_types, schema, use_standard_tags)
     97         if schema is not None:
     98             self._schema = schema

/usr/local/Caskroom/miniconda/base/envs/featuretools/lib/python3.8/site-packages/woodwork/table_accessor.py in _validate_accessor_params(dataframe, index, time_index, logical_types, schema, use_standard_tags)
    877         # We ignore these parameters if a schema is passed
    878         if index is not None:
--> 879             _check_index(dataframe, index)
    880         if logical_types:
    881             _check_logical_types(dataframe.columns, logical_types)

/usr/local/Caskroom/miniconda/base/envs/featuretools/lib/python3.8/site-packages/woodwork/table_accessor.py in _check_index(dataframe, index)
    903         # User specifies an index that is in the dataframe but not unique
    904         # Does not check for Dask as Dask does not support is_unique
--> 905         raise IndexError('Index column must be unique')
    906
    907

IndexError: Index column must be unique

To fix the above error, you can do one of the following solutions:

Solution #1 - You can create a unique index on your Dataframe.

[73]:
product_df = pd.DataFrame({"id": [1, 2, 3, 4, 5], "rating": [3.5, 4.0, 4.5, 1.5, 5.0]})
product_df
[73]:
id rating
0 1 3.5
1 2 4.0
2 3 4.5
3 4 1.5
4 5 5.0

Notice how we now have a unique index column called id.

[74]:
es = es.add_dataframe(dataframe_name="products", dataframe=product_df, index="id")
es
[74]:
Entityset: transactions
  DataFrames:
    transactions [Rows: 500, Columns: 6]
    products [Rows: 5, Columns: 2]
    sessions [Rows: 35, Columns: 5]
    customers [Rows: 5, Columns: 5]
  Relationships:
    transactions.product_id -> products.product_id
    transactions.session_id -> sessions.session_id
    sessions.customer_id -> customers.customer_id

As seen above, we can now create our DataFrame for our EntitySet without an error by creating a unique index in our Dataframe.

Solution #2 - 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.

[75]:
product_df = pd.DataFrame({"id": [1, 2, 3, 4, 4], "rating": [3.5, 4.0, 4.5, 1.5, 5.0]})

es = ft.EntitySet(id="product_data")
es = es.add_dataframe(
    dataframe_name="products", dataframe=product_df, index="product_id", make_index=True
)

es["products"]
[75]:
product_id id rating
0 0 1 3.5
1 1 2 4.0
2 2 3 4.5
3 3 4 1.5
4 4 4 5.0

As seen above, we created our dataframe for our EntitySet without an error using the make_index argument.

Why am I getting the following warning ‘Using training_window but last_time_index is not set’?#

If you are using a training window, and you haven’t set a last_time_index for your dataframe, you will get this warning. The training window attribute in Featuretools limits the amount of past data that can be used while calculating a particular feature vector.

You can add the last_time_index to all dataframes automatically by calling your_entityset.add_last_time_indexes() after you create your EntitySet. This will remove the warning.

[76]:
es = ft.demo.load_mock_customer(return_entityset=True)
es.add_last_time_indexes()

Now we can run DFS without getting the warning.

[77]:
cutoff_times = pd.DataFrame()
cutoff_times["customer_id"] = [1, 2, 3, 1]
cutoff_times["time"] = pd.to_datetime(
    ["2014-1-1 04:00", "2014-1-1 05:00", "2014-1-1 06:00", "2014-1-1 08:00"]
)
cutoff_times["label"] = [True, True, False, True]

feature_matrix, feature_defs = ft.dfs(
    entityset=es,
    target_dataframe_name="customers",
    cutoff_time=cutoff_times,
    cutoff_time_in_index=True,
    training_window="1 hour",
)

last_time_index vs. time_index#

  • The time_index is when the instance was first known.

  • The last_time_index is when the instance appears for the last time.

  • For example, a customer’s session has multiple transactions which can happen at different points in time. If we are trying to count the number of sessions a user has in a given time period, we often want to count all the sessions that had any transaction during the training window. To accomplish this, we need to not only know when a session starts (time_index), but also when it ends (last_time_index). The last time that an instance appears in the data is stored as the last_time_index of a dataframe.

  • Once the last_time_index has been set, Featuretools will check to see if the last_time_index is after the start of the training window. That, combined with the cutoff time, allows DFS to discover which data is relevant for a given training window.

Why am I getting errors with Featuretools on Google Colab?#

Google Colab, by default, has Featuretools 0.4.1 installed. You may run into issues following our newest guides, or latest documentation while using an older version of Featuretools. Therefore, we suggest you upgrade to the latest featuretools version by doing the following in your notebook in Google Colab:

!pip install -U featuretools

You may need to Restart the runtime by doing Runtime -> Restart Runtime. You can check latest Featuretools version by doing following:

import featuretools as ft
print(ft.__version__)

You should see a version greater than 0.4.1