issue on clv_mle/trainer/btyd.py monetary_dnn

Hi,
I am trying to reproduce the google clv prediction documentation. I see in this file within load data function, there is summary dataframe and actual dataframe which is used for calculating error.
# - summary has aggregated values before the threshold date
# - actual_df has values of the overall period.

summary = df_ft[['customer_id', 'frequency_btyd', 'recency', 'T',
'monetary_btyd']]
#[END prob_selec]
summary.columns = ['customer_id', 'frequency', 'recency', 'T',
'monetary_value']
summary = summary.set_index('customer_id')

# additional columns needed for calculating error
actual_df = df_ft[['customer_id', 'frequency_btyd', 'monetary_dnn',
'target_monetary']]

I am confused by 'monetary_dnn' in actual dataframe, is it used correctly here?

monetary_btyd: The average of all orders' monetary values for each customer during the features period.
monetary_dnn: The sum of all orders' monetary values per customer during the features period