Question related to CLV quickstart

[jacktang]

Hello,

I am learning CLV tutorial (https://www.pymc-marketing.io/en/stable/notebooks/clv/clv_quickstart.html), and I am confused about below code:

customer_1516_history = pd.DataFrame(
    dict(
        customer_id=np.arange(10),
        frequency=np.full(10, customer_1516["frequency"], dtype="int"),
        recency=np.full(10, customer_1516["recency"]),
        T=(np.arange(0, 10) + customer_1516["recency"]).astype("int"),
    )
)

Why the customer_id is not customer_1516['customer_id']. When I set it to customer_1516['customer_id'], the code p_alive = bgm.expected_probability_alive(data=customer_1516_history) throws duplicated customer_id error. So how can I understand the customer_id should be unique in customer history?

[jacktang]

I guess the customer_id in dict should be the transaction_id(tx_id)?

[ColtAllen]

Hey @jacktang,

customer_1516 is just a single row. If creating a DataFrame from a dict with customer_id=customer1516['customer_id'], Pandas will auto-replicate it so it's the same length as the other columns, hence the duplicate error you're getting.

Unique transaction IDs could work in this case, but these are not provided in the raw data. Hope this helps.