13. Create Feature List

Create a feature list¶

A feature list is an essential component in machine learning, comprising a collection of features that are used to train models. Let's compile a feature list using some of the features we've created.

For additional features:

• Visit the 'Learn by example' section for a variety of features tailored to different entities and signals.
• Check out the 'Bring Your Own Transformer' tutorials to learn about integrating Large Language Models (LLMs) within the FeatureByte ecosystem.

For those in an enterprise setting, explore 'Ideate Features with FeatureByte Copilot' which adopts an agentic approach to ideate features tailored to your use case.

In [1]:

import featurebyte as fb

# Set your profile to the tutorial environment
fb.use_profile("tutorial")

catalog_name = "Grocery Dataset Tutorial"
catalog = fb.Catalog.activate(catalog_name)

import featurebyte as fb # Set your profile to the tutorial environment fb.use_profile("tutorial") catalog_name = "Grocery Dataset Tutorial" catalog = fb.Catalog.activate(catalog_name)

10:56:48 | WARNING  | Service endpoint is inaccessible: http://featurebyte-server:8088/
10:56:48 | INFO     | Using profile: tutorial
10:56:48 | INFO     | Using configuration file at: /Users/gxav/.featurebyte/config.yaml
10:56:48 | INFO     | Active profile: tutorial (https://tutorials.featurebyte.com/api/v1)
10:56:48 | INFO     | SDK version: 2.0.1.dev67
10:56:48 | INFO     | No catalog activated.
10:56:48 | INFO     | Catalog activated: Grocery Dataset Tutorial
16:11:44 | INFO     | Using configuration file at: /Users/gxav/.featurebyte/config.yaml
16:11:44 | INFO     | Active profile: tutorial (https://tutorials.featurebyte.com/api/v1)
16:11:44 | WARNING  | Remote SDK version (1.1.0.dev7) is different from local (1.1.0.dev1). Update local SDK to avoid unexpected behavior.
16:11:44 | INFO     | No catalog activated.
16:11:44 | INFO     | Catalog activated: Grocery Dataset Tutorial

List all features we created so far¶

In [2]:

catalog.list_features()

catalog.list_features()

Out[2]:

	id	name	dtype	readiness	online_enabled	tables	primary_tables	entities	primary_entities	created_at
0	6762395b79a148e88cc4a645	CUSTOMER_Mean_vector_of_item_product_ProductGr...	FLOAT	DRAFT	False	[GROCERYINVOICE, INVOICEITEMS, GROCERYPRODUCT]	[GROCERYINVOICE, INVOICEITEMS]	[customer]	[customer]	2024-12-18T02:55:46.148000
1	676238eeb0d4bd308cd86255	CUSTOMER_vs_OVERALL_item_TotalCost_across_prod...	FLOAT	DRAFT	False	[GROCERYINVOICE, INVOICEITEMS, GROCERYPRODUCT]	[INVOICEITEMS]	[customer]	[customer]	2024-12-18T02:52:47.291000
2	676238a93fe253046e0e68bb	CUSTOMER_Latest_invoice_Amount_Z_Score_to_invo...	FLOAT	DRAFT	False	[GROCERYINVOICE]	[GROCERYINVOICE]	[customer]	[customer]	2024-12-18T02:51:29.984000
3	676237a960842f89380978dd	CUSTOMER_Std_of_invoice_Amount_28d	FLOAT	DRAFT	False	[GROCERYINVOICE]	[GROCERYINVOICE]	[customer]	[customer]	2024-12-18T02:47:22.367000
4	676237a960842f89380978dc	CUSTOMER_Std_of_invoice_Amount_14d	FLOAT	DRAFT	False	[GROCERYINVOICE]	[GROCERYINVOICE]	[customer]	[customer]	2024-12-18T02:47:21.903000
5	676237a960842f89380978db	CUSTOMER_Avg_of_invoice_Amount_28d	FLOAT	DRAFT	False	[GROCERYINVOICE]	[GROCERYINVOICE]	[customer]	[customer]	2024-12-18T02:47:21.471000
6	676237a960842f89380978da	CUSTOMER_Avg_of_invoice_Amount_14d	FLOAT	DRAFT	False	[GROCERYINVOICE]	[GROCERYINVOICE]	[customer]	[customer]	2024-12-18T02:47:21.022000
7	676237a960842f89380978d9	CUSTOMER_Count_of_invoice_28d	INT	DRAFT	False	[GROCERYINVOICE]	[GROCERYINVOICE]	[customer]	[customer]	2024-12-18T02:47:20.613000
8	676237a960842f89380978d8	CUSTOMER_Count_of_invoice_14d	INT	DRAFT	False	[GROCERYINVOICE]	[GROCERYINVOICE]	[customer]	[customer]	2024-12-18T02:47:20.247000
9	676237a960842f89380978d7	CUSTOMER_Latest_invoice_Amount	FLOAT	DRAFT	False	[GROCERYINVOICE]	[GROCERYINVOICE]	[customer]	[customer]	2024-12-18T02:47:19.766000
10	676237a960842f89380978d3	CUSTOMER_x_PRODUCTGROUP_Sum_of_item_TotalCost_28d	FLOAT	DRAFT	False	[GROCERYINVOICE, INVOICEITEMS, GROCERYPRODUCT]	[INVOICEITEMS]	[customer, productgroup]	[customer, productgroup]	2024-12-18T02:47:19.349000
11	676237a960842f89380978d2	CUSTOMER_x_PRODUCTGROUP_Sum_of_item_TotalCost_14d	FLOAT	DRAFT	False	[GROCERYINVOICE, INVOICEITEMS, GROCERYPRODUCT]	[INVOICEITEMS]	[customer, productgroup]	[customer, productgroup]	2024-12-18T02:47:18.850000
12	676237a960842f89380978d6	CUSTOMER_x_PRODUCTGROUP_Time_Since_Latest_Time...	FLOAT	DRAFT	False	[GROCERYINVOICE, INVOICEITEMS, GROCERYPRODUCT]	[INVOICEITEMS]	[customer, productgroup]	[customer, productgroup]	2024-12-18T02:47:18.197000
13	67623766043662b29958c254	CUSTOMER_Age_band	VARCHAR	DRAFT	False	[GROCERYCUSTOMER]	[GROCERYCUSTOMER]	[customer]	[customer]	2024-12-18T02:46:11.067000
14	67623766043662b29958c24a	CUSTOMER_Age	INT	DRAFT	False	[GROCERYCUSTOMER]	[GROCERYCUSTOMER]	[customer]	[customer]	2024-12-18T02:46:04.282000

Get features from catalog¶

In [3]:

customer_age_band = catalog.get_feature("CUSTOMER_Age_band")
customer_latest_invoice_amount = catalog.get_feature("CUSTOMER_Latest_invoice_Amount")
customer_count_of_invoice_14d = catalog.get_feature("CUSTOMER_Count_of_invoice_14d")
customer_avg_of_invoice_amount_14d = catalog.get_feature("CUSTOMER_Avg_of_invoice_Amount_14d")
customer_std_of_invoice_amount_14d = catalog.get_feature("CUSTOMER_Std_of_invoice_Amount_14d")
customer_latest_invoice_amount_Z_score_to_invoice_amount_28d = catalog.get_feature(
    "CUSTOMER_Latest_invoice_Amount_Z_Score_to_invoice_Amount_28d"
)
customer_vs_overall_item_totalcost_across_product_productgroups_26w = catalog.get_feature(
    "CUSTOMER_vs_OVERALL_item_TotalCost_across_product_ProductGroups_26w"
)
customer_x_productgroup_sum_of_item_totalcost_14d = \
catalog.get_feature("CUSTOMER_x_PRODUCTGROUP_Sum_of_item_TotalCost_14d")
customer_x_productgroup_time_since_latest_timestamp = \
catalog.get_feature("CUSTOMER_x_PRODUCTGROUP_Time_Since_Latest_Timestamp")

customer_age_band = catalog.get_feature("CUSTOMER_Age_band") customer_latest_invoice_amount = catalog.get_feature("CUSTOMER_Latest_invoice_Amount") customer_count_of_invoice_14d = catalog.get_feature("CUSTOMER_Count_of_invoice_14d") customer_avg_of_invoice_amount_14d = catalog.get_feature("CUSTOMER_Avg_of_invoice_Amount_14d") customer_std_of_invoice_amount_14d = catalog.get_feature("CUSTOMER_Std_of_invoice_Amount_14d") customer_latest_invoice_amount_Z_score_to_invoice_amount_28d = catalog.get_feature( "CUSTOMER_Latest_invoice_Amount_Z_Score_to_invoice_Amount_28d" ) customer_vs_overall_item_totalcost_across_product_productgroups_26w = catalog.get_feature( "CUSTOMER_vs_OVERALL_item_TotalCost_across_product_ProductGroups_26w" ) customer_x_productgroup_sum_of_item_totalcost_14d = \ catalog.get_feature("CUSTOMER_x_PRODUCTGROUP_Sum_of_item_TotalCost_14d") customer_x_productgroup_time_since_latest_timestamp = \ catalog.get_feature("CUSTOMER_x_PRODUCTGROUP_Time_Since_Latest_Timestamp")

Create feature list¶

In [4]:

simple_feature_list = fb.FeatureList(
    [
        customer_age_band,
        customer_latest_invoice_amount,
        customer_count_of_invoice_14d,
        customer_avg_of_invoice_amount_14d,
        customer_std_of_invoice_amount_14d,
        customer_latest_invoice_amount_Z_score_to_invoice_amount_28d,
        customer_vs_overall_item_totalcost_across_product_productgroups_26w,
        customer_x_productgroup_sum_of_item_totalcost_14d,
        customer_x_productgroup_time_since_latest_timestamp,
    ],
    name="Customer x ProductGroup Simple FeatureList",
)

simple_feature_list = fb.FeatureList( [ customer_age_band, customer_latest_invoice_amount, customer_count_of_invoice_14d, customer_avg_of_invoice_amount_14d, customer_std_of_invoice_amount_14d, customer_latest_invoice_amount_Z_score_to_invoice_amount_28d, customer_vs_overall_item_totalcost_across_product_productgroups_26w, customer_x_productgroup_sum_of_item_totalcost_14d, customer_x_productgroup_time_since_latest_timestamp, ], name="Customer x ProductGroup Simple FeatureList", )

Preview feature list¶

In [5]:

# Check the primary entity of the feature list
simple_feature_list.primary_entity

# Check the primary entity of the feature list simple_feature_list.primary_entity

Out[5]:

[<featurebyte.api.entity.Entity at 0x316c6dc60>
 {
   'name': 'customer',
   'created_at': '2024-12-18T02:40:54.536000',
   'updated_at': '2024-12-18T02:40:58.932000',
   'description': None,
   'serving_names': [
     'GROCERYCUSTOMERGUID'
   ],
   'catalog_name': 'Grocery Dataset Tutorial'
 },
 <featurebyte.api.entity.Entity at 0x1687c8d60>
 {
   'name': 'productgroup',
   'created_at': '2024-12-18T02:40:55.766000',
   'updated_at': '2024-12-18T02:41:00.663000',
   'description': None,
   'serving_names': [
     'PRODUCTGROUP'
   ],
   'catalog_name': 'Grocery Dataset Tutorial'
 }]

In [6]:

# Get observation table: 'Preview Table with 10 items'
preview_table = catalog.get_observation_table("Preview Table with 10 items")

# Get observation table: 'Preview Table with 10 items' preview_table = catalog.get_observation_table("Preview Table with 10 items")

In [7]:

# Preview simple_feature_list
simple_feature_list.preview(preview_table)

# Preview simple_feature_list simple_feature_list.preview(preview_table)

Out[7]:

	POINT_IN_TIME	GROCERYINVOICEITEMGUID	CUSTOMER_Age_band	CUSTOMER_Latest_invoice_Amount	CUSTOMER_Count_of_invoice_14d	CUSTOMER_Avg_of_invoice_Amount_14d	CUSTOMER_Std_of_invoice_Amount_14d	CUSTOMER_Latest_invoice_Amount_Z_Score_to_invoice_Amount_28d	CUSTOMER_vs_OVERALL_item_TotalCost_across_product_ProductGroups_26w	CUSTOMER_x_PRODUCTGROUP_Sum_of_item_TotalCost_14d	CUSTOMER_x_PRODUCTGROUP_Time_Since_Latest_Timestamp
0	2022-07-23 17:33:25	849454c5-6640-419d-871d-1f0895a1c3b4	35-39	4.64	1.0	4.640000	0.000000	NaN	0.668405	NaN	813.648056
1	2023-01-06 08:49:29	dad86e21-3af4-4b5b-8058-60f946b6dac5	50-54	16.51	7.0	10.618571	3.572868	0.016777	0.640181	NaN	596.949444
2	2022-09-07 09:16:15	53eac49e-7ecd-4cb6-a1c0-38e3900efd7a	80-84	48.67	5.0	41.372000	7.211342	0.901901	0.747673	18.05	69.813889
3	2023-05-21 13:59:44	989643fe-1377-4f5f-8f38-f349a611da0c	65-69	4.79	4.0	22.697500	15.060992	-1.404144	0.722324	4.27	171.889444
4	2023-06-06 16:13:44	bdc6b6bb-a23a-48c8-bb3a-622d9161c0e8	45-49	21.27	NaN	NaN	NaN	1.008156	0.803588	NaN	1007.838611
5	2022-08-22 06:10:42	9267d0dd-9685-4667-8f06-1761abe73c4d	25-29	7.77	2.0	34.075000	26.305000	-0.820547	0.809058	5.99	181.691667
6	2023-03-07 17:53:55	cb478a4e-9266-4523-8ee0-e205881cc5f5	20-24	24.21	NaN	NaN	NaN	NaN	0.677664	NaN	2761.916944
7	2022-12-30 16:42:05	4dd1487a-0379-4eef-b200-97ac1bb1164f	65-69	27.73	8.0	16.381250	9.945394	1.340493	0.800607	0.89	292.786944
8	2022-09-23 18:09:33	459a1b6e-1239-46d1-9e40-539c7e895483	45-49	41.98	2.0	65.835000	23.855000	-0.181665	0.781515	2.69	217.736944
9	2023-03-21 13:49:48	84225bbb-adb8-451e-98c0-897c83c2fad9	65-69	21.90	2.0	19.690000	2.210000	-0.063562	0.803208	7.67	259.306111

Save feature list¶

In [8]:

# Save feature list
simple_feature_list.save()
# Add description
simple_feature_list.update_description("Simple feature list for the customer x productgroup engagement")

# Save feature list simple_feature_list.save() # Add description simple_feature_list.update_description("Simple feature list for the customer x productgroup engagement")

Done! |████████████████████████████████████████| 100% in 6.1s (0.17%/s)         
Done! |████████████████████████████████████████| 100% in 6.1s (0.17%/s)         
Loading Feature(s) |████████████████████████████████████████| 9/9 [100%] in 0.5s
Done! |████████████████████████████████████████| 100% in 6.1s (0.17%/s)         
Done! |████████████████████████████████████████| 100% in 6.1s (0.17%/s)         
Loading Feature(s) |████████████████████████████████████████| 9/9 [100%] in 0.4s

Concepts in this tutorial¶

• More on feature lists

SDK Reference for¶

• Feature List
• FeatureList.save()