11. Create Calendar Window Aggregates from Time Series

Creating Features from Monthly Installments¶

In this step, we define 9 features by aggregating monthly installment records from the PRIOR_APPLICATIONS table within 12-month and 24-month calendar windows, after applying a filter on the INSTALLMENT_STATUS column.

1. Statistical Aggregations (Max, Min, Avg)¶

These features summarize consumer installment records using basic statistical functions:

Maximum (Max)¶

CLIENT_Max_of_Consumer_installments_records_Payment_Delays_24cMo
CLIENT_Max_of_Consumer_installments_records_Consumer_Loan_AMT_APPLICATIONs_24cMo

Minimum (Min)¶

CLIENT_Min_of_Consumer_installments_records_Consumer_Loan_AMT_CREDIT_To_AMT_APPLICATIONs_24cMo
CLIENT_Min_of_Consumer_installments_records_AMT_PAYMENTs_12cMo

Average (Avg)¶

CLIENT_Avg_of_Consumer_installments_records_AMT_PAYMENTs_24cMo
CLIENT_Avg_of_Late_Status_Consumer_installments_records_AMT_PAYMENTs_24cMo

2. Temporal Mean Derived from Two Features¶

This feature computes a temporal mean by dividing the sum of annuities by the number of unique installment dates:

CLIENT_Temporal_Mean_of_Sums_of_Consumer_installments_records_Consumer_Loan_AMT_ANNUITYs_24cMo

3. Aggregations Across `INSTALLMENT_STATUS` or `YIELD_GROUP`¶

These features summarize installment payments across different categories:

CLIENT_Consumer_installments_records_AMT_PAYMENTs_by_Consumer_installments_record_INSTALLMENT_STATUS_6cMo
CLIENT_Count_of_Consumer_installments_records_by_Consumer_Loan_YIELD_GROUP_24cMo

Activate catalog¶

See SDK reference for catalogs

In [1]:

            
                Copied!
                
import featurebyte as fb

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

catalog_name = "Credit Default Dataset SDK Tutorial"
catalog = fb.Catalog.activate(catalog_name)
import featurebyte as fb

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

catalog_name = "Credit Default Dataset SDK Tutorial"
catalog = fb.Catalog.activate(catalog_name)

16:43:19 | WARNING  | Service endpoint is inaccessible: http://featurebyte-server:8088/
16:43:19 | INFO     | Using profile: tutorial
16:43:20 | INFO     | Using configuration file at: /Users/gxav/.featurebyte/config.yaml
16:43:20 | INFO     | Active profile: tutorial (https://tutorials.featurebyte.com/api/v1)
16:43:20 | INFO     | SDK version: 2.1.0.dev113
16:43:20 | INFO     | No catalog activated.
16:43:20 | INFO     | Catalog activated: Credit Default Dataset SDK Tutorial

Get view from table¶

See SDK reference for views

In [2]:

            
                Copied!
                
# Get view from CONSUMER_LOAN_STATUS scd table.
consumer_loan_status_view = catalog.get_view("CONSUMER_LOAN_STATUS")
# Get view from CONSUMER_LOAN_STATUS scd table.
consumer_loan_status_view = catalog.get_view("CONSUMER_LOAN_STATUS")

In [3]:

            
                Copied!
                
# Get view from CONSUMER_INSTALLMENTS time series table.
consumer_installments_view = catalog.get_view("CONSUMER_INSTALLMENTS")
# Get view from CONSUMER_INSTALLMENTS time series table.
consumer_installments_view = catalog.get_view("CONSUMER_INSTALLMENTS")

Create ratio and time delta columns¶

See SDK reference for view columns

In [4]:

            
                Copied!
                
consumer_loan_status_view["AMT_CREDIT To AMT_APPLICATION"] = (
    consumer_loan_status_view["AMT_CREDIT"]
    / consumer_loan_status_view["AMT_APPLICATION"]
)
consumer_loan_status_view["AMT_CREDIT To AMT_APPLICATION"] = (
    consumer_loan_status_view["AMT_CREDIT"]
    / consumer_loan_status_view["AMT_APPLICATION"]
)

In [5]:

            
                Copied!
                
consumer_installments_view["Payment Delay"] = (
    consumer_installments_view["ACTUAL_INSTALLMENT_DATE"]
    - consumer_installments_view["SCHEDULED_INSTALLMENT_DATE"]
).dt.day
consumer_installments_view["Payment Delay"] = (
    consumer_installments_view["ACTUAL_INSTALLMENT_DATE"]
    - consumer_installments_view["SCHEDULED_INSTALLMENT_DATE"]
).dt.day

Join views¶

See SDK reference for views

In [6]:

            
                Copied!
                
# Join CONSUMER_LOAN_STATUS view to CONSUMER_INSTALLMENTS view.
consumer_installments_view = consumer_installments_view.join(
    consumer_loan_status_view, rprefix="Consumer Loan_"
)
# Join CONSUMER_LOAN_STATUS view to CONSUMER_INSTALLMENTS view.
consumer_installments_view = consumer_installments_view.join(
    consumer_loan_status_view, rprefix="Consumer Loan_"
)

Filter View¶

Read on view subsetting

In [7]:

            
                Copied!
                
# Filter consumer_installments_view with INSTALLMENT_STATUS equal to Late.
cond = consumer_installments_view["INSTALLMENT_STATUS"] == "Late"
consumer_installments_late_status_view = consumer_installments_view[cond]
# Filter consumer_installments_view with INSTALLMENT_STATUS equal to Late.
cond = consumer_installments_view["INSTALLMENT_STATUS"] == "Late"
consumer_installments_late_status_view = consumer_installments_view[cond]

Declare simple stats from CONSUMER_INSTALLMENTS¶

See SDK reference for features

See SDK reference to groupby a view

See SDK reference to do aggregation over time

See SDK reference for feature group

See list of aggregation operations

In [8]:

            
                Copied!
                
# Group CONSUMER_INSTALLMENTS view by Client entity (Consumer Loan_CLIENT_ID).
consumer_installments_view_by_client = consumer_installments_view.groupby(["Consumer Loan_CLIENT_ID"])
# Group CONSUMER_INSTALLMENTS view by Client entity (Consumer Loan_CLIENT_ID).
consumer_installments_view_by_client = consumer_installments_view.groupby(["Consumer Loan_CLIENT_ID"])

In [9]:

            
                Copied!
                
                    
                    
                
                

        
# Get Min of AMT_PAYMENT for the Client over time.
client_min_of_consumer_installments_records_amt_payments_12cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "AMT_PAYMENT",
        method="min",
        feature_names=[
            "CLIENT_Min_of_Consumer_installments_records_AMT_PAYMENTs_12cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=12)],
    )["CLIENT_Min_of_Consumer_installments_records_AMT_PAYMENTs_12cMo"]
)
# Get Min of AMT_PAYMENT for the Client over time.
client_min_of_consumer_installments_records_amt_payments_12cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "AMT_PAYMENT",
        method="min",
        feature_names=[
            "CLIENT_Min_of_Consumer_installments_records_AMT_PAYMENTs_12cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=12)],
    )["CLIENT_Min_of_Consumer_installments_records_AMT_PAYMENTs_12cMo"]
)

In [10]:

            
                Copied!
                
                    
                    
                
                

        
# Get Avg of AMT_PAYMENT for the Client over time.
client_avg_of_consumer_installments_records_amt_payments_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "AMT_PAYMENT",
        method="avg",
        feature_names=[
            "CLIENT_Avg_of_Consumer_installments_records_AMT_PAYMENTs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Avg_of_Consumer_installments_records_AMT_PAYMENTs_24cMo"]
)
# Get Avg of AMT_PAYMENT for the Client over time.
client_avg_of_consumer_installments_records_amt_payments_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "AMT_PAYMENT",
        method="avg",
        feature_names=[
            "CLIENT_Avg_of_Consumer_installments_records_AMT_PAYMENTs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Avg_of_Consumer_installments_records_AMT_PAYMENTs_24cMo"]
)

In [11]:

            
                Copied!
                
                    
                    
                
                

        
# Get Max of Payment Delay for the Client over time.
client_max_of_consumer_installments_records_payment_delays_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "Payment Delay",
        method="max",
        feature_names=[
            "CLIENT_Max_of_Consumer_installments_records_Payment_Delays_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Max_of_Consumer_installments_records_Payment_Delays_24cMo"]
)
# Get Max of Payment Delay for the Client over time.
client_max_of_consumer_installments_records_payment_delays_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "Payment Delay",
        method="max",
        feature_names=[
            "CLIENT_Max_of_Consumer_installments_records_Payment_Delays_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Max_of_Consumer_installments_records_Payment_Delays_24cMo"]
)

In [12]:

            
                Copied!
                
                    
                    
                
                

        
# Get Max of Consumer Loan_AMT_APPLICATION for the Client over time.
client_max_of_consumer_installments_records_consumer_loan_amt_applications_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "Consumer Loan_AMT_APPLICATION",
        method="max",
        feature_names=[
            "CLIENT_Max_of_Consumer_installments_records_Consumer_Loan_AMT_APPLICATIONs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Max_of_Consumer_installments_records_Consumer_Loan_AMT_APPLICATIONs_24cMo"]
)
# Get Max of Consumer Loan_AMT_APPLICATION for the Client over time.
client_max_of_consumer_installments_records_consumer_loan_amt_applications_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "Consumer Loan_AMT_APPLICATION",
        method="max",
        feature_names=[
            "CLIENT_Max_of_Consumer_installments_records_Consumer_Loan_AMT_APPLICATIONs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Max_of_Consumer_installments_records_Consumer_Loan_AMT_APPLICATIONs_24cMo"]
)

In [13]:

            
                Copied!
                
                    
                    
                
                

        
# Get Min of Consumer Loan_AMT_CREDIT To AMT_APPLICATION for the Client over time.
client_min_of_consumer_installments_records_consumer_loan_amt_credit_to_amt_applications_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "Consumer Loan_AMT_CREDIT To AMT_APPLICATION",
        method="min",
        feature_names=[
            "CLIENT_Min_of_Consumer_installments_records_Consumer_Loan_AMT_CREDIT_To_AMT_APPLICATIONs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Min_of_Consumer_installments_records_Consumer_Loan_AMT_CREDIT_To_AMT_APPLICATIONs_24cMo"]
)
# Get Min of Consumer Loan_AMT_CREDIT To AMT_APPLICATION for the Client over time.
client_min_of_consumer_installments_records_consumer_loan_amt_credit_to_amt_applications_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "Consumer Loan_AMT_CREDIT To AMT_APPLICATION",
        method="min",
        feature_names=[
            "CLIENT_Min_of_Consumer_installments_records_Consumer_Loan_AMT_CREDIT_To_AMT_APPLICATIONs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Min_of_Consumer_installments_records_Consumer_Loan_AMT_CREDIT_To_AMT_APPLICATIONs_24cMo"]
)

Declare temporal mean from two features¶

In [14]:

            
                Copied!
                
                    
                    
                
                

        
# Get Unique Count of ACTUAL_INSTALLMENT_DATE for the Client over time.
client_temporary_unique_count_of_consumer_installments_records_actual_installment_dates_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "ACTUAL_INSTALLMENT_DATE",
        method="count_distinct",
        feature_names=[
            "CLIENT_Temporary_Unique_Count_of_Consumer_installments_records_ACTUAL_INSTALLMENT_DATEs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Temporary_Unique_Count_of_Consumer_installments_records_ACTUAL_INSTALLMENT_DATEs_24cMo"]
)
# Get Unique Count of ACTUAL_INSTALLMENT_DATE for the Client over time.
client_temporary_unique_count_of_consumer_installments_records_actual_installment_dates_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "ACTUAL_INSTALLMENT_DATE",
        method="count_distinct",
        feature_names=[
            "CLIENT_Temporary_Unique_Count_of_Consumer_installments_records_ACTUAL_INSTALLMENT_DATEs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Temporary_Unique_Count_of_Consumer_installments_records_ACTUAL_INSTALLMENT_DATEs_24cMo"]
)

In [15]:

            
                Copied!
                
                    
                    
                
                

        
# Get Sum of Consumer Loan_AMT_ANNUITY for the Client over time.
client_temporary_sum_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "Consumer Loan_AMT_ANNUITY",
        method="sum",
        feature_names=[
            "CLIENT_Temporary_Sum_of_Consumer_installments_records_Consumer_Loan_AMT_ANNUITYs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Temporary_Sum_of_Consumer_installments_records_Consumer_Loan_AMT_ANNUITYs_24cMo"]
)
# Get Sum of Consumer Loan_AMT_ANNUITY for the Client over time.
client_temporary_sum_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo = (
    consumer_installments_view_by_client.aggregate_over(
        "Consumer Loan_AMT_ANNUITY",
        method="sum",
        feature_names=[
            "CLIENT_Temporary_Sum_of_Consumer_installments_records_Consumer_Loan_AMT_ANNUITYs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Temporary_Sum_of_Consumer_installments_records_Consumer_Loan_AMT_ANNUITYs_24cMo"]
)

In [16]:

            
                Copied!
                
                    
                    
                
                

        
# Derive Average of Sums
client_temporal_mean_of_sums_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo = (
    client_temporary_sum_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo
    / client_temporary_unique_count_of_consumer_installments_records_actual_installment_dates_24cmo
)

# Give a name to new feature
client_temporal_mean_of_sums_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo.name = (
    "CLIENT_Temporal_Mean_of_Sums_of_Consumer_installments_records_Consumer_Loan_AMT_ANNUITYs_24cMo"
)
# Derive Average of Sums
client_temporal_mean_of_sums_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo = (
    client_temporary_sum_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo
    / client_temporary_unique_count_of_consumer_installments_records_actual_installment_dates_24cmo
)

# Give a name to new feature
client_temporal_mean_of_sums_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo.name = (
    "CLIENT_Temporal_Mean_of_Sums_of_Consumer_installments_records_Consumer_Loan_AMT_ANNUITYs_24cMo"
)

Do window aggregation across different INSTALLMENT_STATUS¶

In [17]:

            
                Copied!
                
                    
                    
                
                

        
# Group CONSUMER_INSTALLMENTS view by Client entity (Consumer Loan_CLIENT_ID) across different INSTALLMENT_STATUS.
consumer_installments_view_by_client_across_installment_statuss = (
    consumer_installments_view.groupby(
        ["Consumer Loan_CLIENT_ID"], category="INSTALLMENT_STATUS"
    )
)
# Group CONSUMER_INSTALLMENTS view by Client entity (Consumer Loan_CLIENT_ID) across different INSTALLMENT_STATUS.
consumer_installments_view_by_client_across_installment_statuss = (
    consumer_installments_view.groupby(
        ["Consumer Loan_CLIENT_ID"], category="INSTALLMENT_STATUS"
    )
)

In [18]:

            
                Copied!
                
                    
                    
                
                

        
# Distribution of the total AMT_PAYMENTs of Consumer_installments_records, segmented by
# Consumer_installments_record's INSTALLMENT_STATUS for the Client over time.
client_consumer_installments_records_amt_payments_by_consumer_installments_record_installment_status_6cmo = (
    consumer_installments_view_by_client_across_installment_statuss.aggregate_over(
        "AMT_PAYMENT",
        method="sum",
        feature_names=[
            "CLIENT_Consumer_installments_records_AMT_PAYMENTs_by_Consumer_installments_record_INSTALLMENT_STATUS_6cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=6)],
    )["CLIENT_Consumer_installments_records_AMT_PAYMENTs_by_Consumer_installments_record_INSTALLMENT_STATUS_6cMo"]
)
# Distribution of the total AMT_PAYMENTs of Consumer_installments_records, segmented by
# Consumer_installments_record's INSTALLMENT_STATUS for the Client over time.
client_consumer_installments_records_amt_payments_by_consumer_installments_record_installment_status_6cmo = (
    consumer_installments_view_by_client_across_installment_statuss.aggregate_over(
        "AMT_PAYMENT",
        method="sum",
        feature_names=[
            "CLIENT_Consumer_installments_records_AMT_PAYMENTs_by_Consumer_installments_record_INSTALLMENT_STATUS_6cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=6)],
    )["CLIENT_Consumer_installments_records_AMT_PAYMENTs_by_Consumer_installments_record_INSTALLMENT_STATUS_6cMo"]
)

Do window aggregation across different Loan_YIELD_GROUPs¶

In [19]:

            
                Copied!
                
                    
                    
                
                

        
# Group CONSUMER_INSTALLMENTS view by Client entity (Consumer Loan_CLIENT_ID) across different
# Consumer Loan_YIELD_GROUPs.
consumer_installments_view_by_client_across_consumer_loan_yield_groups = (
    consumer_installments_view.groupby(
        ["Consumer Loan_CLIENT_ID"], category="Consumer Loan_YIELD_GROUP"
    )
)
# Group CONSUMER_INSTALLMENTS view by Client entity (Consumer Loan_CLIENT_ID) across different
# Consumer Loan_YIELD_GROUPs.
consumer_installments_view_by_client_across_consumer_loan_yield_groups = (
    consumer_installments_view.groupby(
        ["Consumer Loan_CLIENT_ID"], category="Consumer Loan_YIELD_GROUP"
    )
)

In [20]:

            
                Copied!
                
                    
                    
                
                

        
# Count Consumer_installments_records across different Consumer Loan_YIELD_GROUPs for the Client
# over time.
client_count_of_consumer_installments_records_by_consumer_loan_yield_group_24cmo = (
    consumer_installments_view_by_client_across_consumer_loan_yield_groups.aggregate_over(
        None,
        method="count",
        feature_names=[
            "CLIENT_Count_of_Consumer_installments_records_by_Consumer_Loan_YIELD_GROUP_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Count_of_Consumer_installments_records_by_Consumer_Loan_YIELD_GROUP_24cMo"]
)
# Count Consumer_installments_records across different Consumer Loan_YIELD_GROUPs for the Client
# over time.
client_count_of_consumer_installments_records_by_consumer_loan_yield_group_24cmo = (
    consumer_installments_view_by_client_across_consumer_loan_yield_groups.aggregate_over(
        None,
        method="count",
        feature_names=[
            "CLIENT_Count_of_Consumer_installments_records_by_Consumer_Loan_YIELD_GROUP_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Count_of_Consumer_installments_records_by_Consumer_Loan_YIELD_GROUP_24cMo"]
)

Do window aggregation from consumer_installments_late_status_view¶

See SDK reference for features

See SDK reference to groupby a view

See SDK reference to do aggregation over time

See SDK reference for feature group

See list of aggregation operations

In [21]:

            
                Copied!
                
# Group consumer_installments_late_status_view view by Client entity (Consumer Loan_CLIENT_ID).
consumer_installments_late_status_view_by_client = (
    consumer_installments_late_status_view.groupby(["Consumer Loan_CLIENT_ID"])
)
# Group consumer_installments_late_status_view view by Client entity (Consumer Loan_CLIENT_ID).
consumer_installments_late_status_view_by_client = (
    consumer_installments_late_status_view.groupby(["Consumer Loan_CLIENT_ID"])
)

In [22]:

            
                Copied!
                
                    
                    
                
                

        
# Get Avg of Late Status AMT_PAYMENT for the Client over time.
client_avg_of_late_status_consumer_installments_records_amt_payments_24cmo = (
    consumer_installments_late_status_view_by_client.aggregate_over(
        "AMT_PAYMENT",
        method="avg",
        feature_names=[
            "CLIENT_Avg_of_Late_Status_Consumer_installments_records_AMT_PAYMENTs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Avg_of_Late_Status_Consumer_installments_records_AMT_PAYMENTs_24cMo"]
)
# Get Avg of Late Status AMT_PAYMENT for the Client over time.
client_avg_of_late_status_consumer_installments_records_amt_payments_24cmo = (
    consumer_installments_late_status_view_by_client.aggregate_over(
        "AMT_PAYMENT",
        method="avg",
        feature_names=[
            "CLIENT_Avg_of_Late_Status_Consumer_installments_records_AMT_PAYMENTs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["CLIENT_Avg_of_Late_Status_Consumer_installments_records_AMT_PAYMENTs_24cMo"]
)

Do window aggregation on all data¶

In [23]:

            
                Copied!
                
# Group CONSUMER_INSTALLMENTS view without any groupby key for aggregates on all data.
consumer_installments_view_by_overall = consumer_installments_view.groupby([])
# Group CONSUMER_INSTALLMENTS view without any groupby key for aggregates on all data.
consumer_installments_view_by_overall = consumer_installments_view.groupby([])

In [24]:

            
                Copied!
                
                    
                    
                
                

        
# Get Avg of AMT_PAYMENT over time.
overall_avg_of_consumer_installments_records_amt_payments_24cmo = (
    consumer_installments_view_by_overall.aggregate_over(
        "AMT_PAYMENT",
        method="avg",
        feature_names=[
            "OVERALL_Avg_of_Consumer_installments_records_AMT_PAYMENTs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["OVERALL_Avg_of_Consumer_installments_records_AMT_PAYMENTs_24cMo"]
)
# Get Avg of AMT_PAYMENT over time.
overall_avg_of_consumer_installments_records_amt_payments_24cmo = (
    consumer_installments_view_by_overall.aggregate_over(
        "AMT_PAYMENT",
        method="avg",
        feature_names=[
            "OVERALL_Avg_of_Consumer_installments_records_AMT_PAYMENTs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["OVERALL_Avg_of_Consumer_installments_records_AMT_PAYMENTs_24cMo"]
)

In [25]:

            
                Copied!
                
# Group consumer_installments_late_status_view view without any groupby key for aggregates on all
# data.
consumer_installments_late_status_view_by_overall = (
    consumer_installments_late_status_view.groupby([])
)
# Group consumer_installments_late_status_view view without any groupby key for aggregates on all
# data.
consumer_installments_late_status_view_by_overall = (
    consumer_installments_late_status_view.groupby([])
)

In [26]:

            
                Copied!
                
                    
                    
                
                

        
# Get Avg of Late Status AMT_PAYMENT over time.
overall_avg_of_late_status_consumer_installments_records_amt_payments_24cmo = (
    consumer_installments_late_status_view_by_overall.aggregate_over(
        "AMT_PAYMENT",
        method="avg",
        feature_names=[
            "OVERALL_Avg_of_Late_Status_Consumer_installments_records_AMT_PAYMENTs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["OVERALL_Avg_of_Late_Status_Consumer_installments_records_AMT_PAYMENTs_24cMo"]
)
# Get Avg of Late Status AMT_PAYMENT over time.
overall_avg_of_late_status_consumer_installments_records_amt_payments_24cmo = (
    consumer_installments_late_status_view_by_overall.aggregate_over(
        "AMT_PAYMENT",
        method="avg",
        feature_names=[
            "OVERALL_Avg_of_Late_Status_Consumer_installments_records_AMT_PAYMENTs_24cMo"
        ],
        windows=[fb.CalendarWindow(unit="MONTH", size=24)],
    )["OVERALL_Avg_of_Late_Status_Consumer_installments_records_AMT_PAYMENTs_24cMo"]
)

Save features¶

See SDK reference to save feature

See SDK reference to save feature group

In [27]:

            
                Copied!
                
                    
                    
                
                

        
fb.FeatureGroup(
    [
        client_max_of_consumer_installments_records_payment_delays_24cmo,
        client_max_of_consumer_installments_records_consumer_loan_amt_applications_24cmo,
        client_min_of_consumer_installments_records_consumer_loan_amt_credit_to_amt_applications_24cmo,
        client_min_of_consumer_installments_records_amt_payments_12cmo,
        client_avg_of_consumer_installments_records_amt_payments_24cmo,
        client_avg_of_late_status_consumer_installments_records_amt_payments_24cmo,
        client_temporal_mean_of_sums_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo,
        client_count_of_consumer_installments_records_by_consumer_loan_yield_group_24cmo,
        client_consumer_installments_records_amt_payments_by_consumer_installments_record_installment_status_6cmo,
    ]
).save()
fb.FeatureGroup(
    [
        client_max_of_consumer_installments_records_payment_delays_24cmo,
        client_max_of_consumer_installments_records_consumer_loan_amt_applications_24cmo,
        client_min_of_consumer_installments_records_consumer_loan_amt_credit_to_amt_applications_24cmo,
        client_min_of_consumer_installments_records_amt_payments_12cmo,
        client_avg_of_consumer_installments_records_amt_payments_24cmo,
        client_avg_of_late_status_consumer_installments_records_amt_payments_24cmo,
        client_temporal_mean_of_sums_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo,
        client_count_of_consumer_installments_records_by_consumer_loan_yield_group_24cmo,
        client_consumer_installments_records_amt_payments_by_consumer_installments_record_installment_status_6cmo,
    ]
).save()

Done! |████████████████████████████████████████| 100% in 9.1s (0.11%/s)         
Done! |████████████████████████████████████████| 100% in 6.1s (0.17%/s)         
Loading Feature(s) |████████████████████████████████████████| 9/9 [100%] in 0.4s

Add description¶

In [28]:

            
                Copied!
                
                    
                    
                
                

        
client_max_of_consumer_installments_records_payment_delays_24cmo.update_description(
    "Max of Consumer_installments_records Payment Delays for the Client "
    "over a 24 calendar months period."
)
client_max_of_consumer_installments_records_consumer_loan_amt_applications_24cmo.update_description(
    "Max of Consumer_installments_records Consumer Loan_AMT_APPLICATIONs "
    "for the Client over a 24 calendar months period."
)
client_min_of_consumer_installments_records_consumer_loan_amt_credit_to_amt_applications_24cmo.update_description(
    "Min of Consumer_installments_records Consumer Loan_AMT_CREDIT To "
    "AMT_APPLICATIONs for the Client over a 24 calendar months period."
)
client_min_of_consumer_installments_records_amt_payments_12cmo.update_description(
    "Min of Consumer_installments_records AMT_PAYMENTs for the Client over "
    "a 12 calendar months period."
)
client_avg_of_consumer_installments_records_amt_payments_24cmo.update_description(
    "Avg of Consumer_installments_records AMT_PAYMENTs for "
    "the Client over a 24 calendar months period."
)
client_avg_of_late_status_consumer_installments_records_amt_payments_24cmo.update_description(
    "Avg of Late Status Consumer_installments_records AMT_PAYMENTs for "
    "the Client over a 24 calendar months period."
)
client_temporal_mean_of_sums_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo.update_description(
    "Temporal Avg of Sum of Consumer_installments_records Consumer "
    "Loan_AMT_ANNUITYs for the Client over a 24 calendar months period."
)
client_count_of_consumer_installments_records_by_consumer_loan_yield_group_24cmo.update_description(
    "Count of Consumer_installments_records across different Consumer "
    "Loan_YIELD_GROUP categories for the Client over a 24 calendar months "
    "period. It provides a detailed breakdown in the form of a dictionary, "
    "where each key represents a unique Consumer Loan_YIELD_GROUP. The "
    "corresponding value for each key is the count within that Consumer "
    "Loan_YIELD_GROUP. This distribution offers insights into the spread "
    "across different Consumer Loan_YIELD_GROUPs for the Client."
)
client_consumer_installments_records_amt_payments_by_consumer_installments_record_installment_status_6cmo.update_description(
    "Distribution of the total AMT_PAYMENTs of "
    "Consumer_installments_records, segmented by "
    "Consumer_installments_record's INSTALLMENT_STATUS for the Client over "
    "a 6 calendar months period. It provides a detailed breakdown in the "
    "form of a dictionary, where each key represents a unique "
    "INSTALLMENT_STATUS. The corresponding value for each key is the "
    "cumulative sum of AMT_PAYMENTs within that INSTALLMENT_STATUS. This "
    "distribution offers insights into the AMT_PAYMENT spread across "
    "different INSTALLMENT_STATUSs for the Client."
)
client_max_of_consumer_installments_records_payment_delays_24cmo.update_description(
    "Max of Consumer_installments_records Payment Delays for the Client "
    "over a 24 calendar months period."
)
client_max_of_consumer_installments_records_consumer_loan_amt_applications_24cmo.update_description(
    "Max of Consumer_installments_records Consumer Loan_AMT_APPLICATIONs "
    "for the Client over a 24 calendar months period."
)
client_min_of_consumer_installments_records_consumer_loan_amt_credit_to_amt_applications_24cmo.update_description(
    "Min of Consumer_installments_records Consumer Loan_AMT_CREDIT To "
    "AMT_APPLICATIONs for the Client over a 24 calendar months period."
)
client_min_of_consumer_installments_records_amt_payments_12cmo.update_description(
    "Min of Consumer_installments_records AMT_PAYMENTs for the Client over "
    "a 12 calendar months period."
)
client_avg_of_consumer_installments_records_amt_payments_24cmo.update_description(
    "Avg of Consumer_installments_records AMT_PAYMENTs for "
    "the Client over a 24 calendar months period."
)
client_avg_of_late_status_consumer_installments_records_amt_payments_24cmo.update_description(
    "Avg of Late Status Consumer_installments_records AMT_PAYMENTs for "
    "the Client over a 24 calendar months period."
)
client_temporal_mean_of_sums_of_consumer_installments_records_consumer_loan_amt_annuitys_24cmo.update_description(
    "Temporal Avg of Sum of Consumer_installments_records Consumer "
    "Loan_AMT_ANNUITYs for the Client over a 24 calendar months period."
)
client_count_of_consumer_installments_records_by_consumer_loan_yield_group_24cmo.update_description(
    "Count of Consumer_installments_records across different Consumer "
    "Loan_YIELD_GROUP categories for the Client over a 24 calendar months "
    "period. It provides a detailed breakdown in the form of a dictionary, "
    "where each key represents a unique Consumer Loan_YIELD_GROUP. The "
    "corresponding value for each key is the count within that Consumer "
    "Loan_YIELD_GROUP. This distribution offers insights into the spread "
    "across different Consumer Loan_YIELD_GROUPs for the Client."
)
client_consumer_installments_records_amt_payments_by_consumer_installments_record_installment_status_6cmo.update_description(
    "Distribution of the total AMT_PAYMENTs of "
    "Consumer_installments_records, segmented by "
    "Consumer_installments_record's INSTALLMENT_STATUS for the Client over "
    "a 6 calendar months period. It provides a detailed breakdown in the "
    "form of a dictionary, where each key represents a unique "
    "INSTALLMENT_STATUS. The corresponding value for each key is the "
    "cumulative sum of AMT_PAYMENTs within that INSTALLMENT_STATUS. This "
    "distribution offers insights into the AMT_PAYMENT spread across "
    "different INSTALLMENT_STATUSs for the Client."
)

In [ ]: