cancel
Showing results for
Did you mean:
Regular Visitor

## GROUP BY with DISTINCT COUNT

I am looking to calculate the unique customer count from the groupby function.

Python code would be :

df = temp_df.groupby(['product','od_pairs']).agg(total_bookings=('booking_no_id', 'count'),
total_teus=('total_teus','sum'),
total_bl=('bl_amount', 'sum'),
total_cost=('cost', 'sum'),
total_customer=('control_party_name', 'nunique'),
)

How do i incorporate distinct count in the groupby function?

summary_table = GROUPBY( export_data, export_data[customer_segment], export_data[product], export_data[main_service], export_data[vessel_direction], export_data[od_pairs],
"booking_count", COUNTX(CURRENTGROUP(), export_data[booking_no_id]),
"total_teus", SUMX(CURRENTGROUP(), export_data[total_teus]),
"total_bl", SUMX(CURRENTGROUP(), export_data[bl_amount]),
"total_cost", SUMX(CURRENTGROUP(), export_data[cost]),
"total_cust", COUNTX(CURRENTGROUP(), DISTINCT(export_data[control_party_name]))

1 ACCEPTED SOLUTION
Super User

Hi @jeanL

I don't believe there's any way to calculate a distinct count using GROUPBY.

Depending where you need to use this table expression, I would recommend rewriting with SUMMARIZECOLUMNS (if you are materializing this table or running as a query) otherwise with SUMMARIZE:

(hopefully I interpreted your intended result correctly):

SUMMARIZECOLUMNS version

``````summary_table =
SUMMARIZECOLUMNS (
export_data[customer_segment],
export_data[product],
export_data[main_service],
export_data[vessel_direction],
export_data[od_pairs],
"booking_count", COUNT ( export_data[booking_no_id] ),
"total_teus", SUM ( export_data[total_teus] ),
"total_bl", SUM ( export_data[bl_amount] ),
"total_cost", SUM ( export_data[cost] ),
"total_cust", DISTINCTCOUNT ( export_data[control_party_name] )
)
``````

``````summary_table =
SUMMARIZE (
export_data,
export_data[customer_segment],
export_data[product],
export_data[main_service],
export_data[vessel_direction],
export_data[od_pairs]
),
"booking_count", CALCULATE ( COUNT ( export_data[booking_no_id] ) ),
"total_teus", CALCULATE ( SUM ( export_data[total_teus] ) ),
"total_bl", CALCULATE ( SUM ( export_data[bl_amount] ) ),
"total_cost", CALCULATE ( SUM ( export_data[cost] ) ),
"total_cust", CALCULATE ( DISTINCTCOUNT ( export_data[control_party_name] ) )
)
``````

Regards,

Owen

Owen Auger

My Blog
2 REPLIES 2
Regular Visitor

Thanks @OwenAuger ! It works.

Super User

Hi @jeanL

I don't believe there's any way to calculate a distinct count using GROUPBY.

Depending where you need to use this table expression, I would recommend rewriting with SUMMARIZECOLUMNS (if you are materializing this table or running as a query) otherwise with SUMMARIZE:

(hopefully I interpreted your intended result correctly):

SUMMARIZECOLUMNS version

``````summary_table =
SUMMARIZECOLUMNS (
export_data[customer_segment],
export_data[product],
export_data[main_service],
export_data[vessel_direction],
export_data[od_pairs],
"booking_count", COUNT ( export_data[booking_no_id] ),
"total_teus", SUM ( export_data[total_teus] ),
"total_bl", SUM ( export_data[bl_amount] ),
"total_cost", SUM ( export_data[cost] ),
"total_cust", DISTINCTCOUNT ( export_data[control_party_name] )
)
``````

``````summary_table =
SUMMARIZE (
export_data,
export_data[customer_segment],
export_data[product],
export_data[main_service],
export_data[vessel_direction],
export_data[od_pairs]
),
"booking_count", CALCULATE ( COUNT ( export_data[booking_no_id] ) ),
"total_teus", CALCULATE ( SUM ( export_data[total_teus] ) ),
"total_bl", CALCULATE ( SUM ( export_data[bl_amount] ) ),
"total_cost", CALCULATE ( SUM ( export_data[cost] ) ),
"total_cust", CALCULATE ( DISTINCTCOUNT ( export_data[control_party_name] ) )
)
``````

Regards,

Owen

Owen Auger

My Blog

Announcements