CODEWITHSUNDEEP
sqlcombinationsdata-analysispresto
Chris90
Chris90

Reputation: 1998

SQL Presto Query - Retrieving all possible combinations of rows?

I am wondering how I can get all possible combinations of a cross joined table to itself?

Sample table looks like 

    DAY   Order  pickup_lat  pickup_long     dropoff_lat dropoff_long  created_time
 1/3/19  234e    32.69        -117.1          32.63      -117.08   3/1/19 19:00
 1/3/19  235d    40.73        -73.98          40.73       -73.99   3/1/19 23:21
 1/3/19  253w    40.76        -73.99          40.76       -73.99   3/1/19 15:26
 2/3/19  231y    36.08        -94.2           36.07       -94.21   3/2/19 0:14
 3/3/19  305g    36.01        -78.92          36.01       -78.95   3/2/19 0:09
 3/3/19  328s    36.76        -119.83         36.74       -119.79  3/2/19 4:33
 3/3/19  286n    35.76        -78.78          35.78       -78.74   3/2/19 0:43

I want to see all possible combinations of orders based on their differences in order created times and distance in miles for both pick-up and drop-off. Is this possible?

I would be using great_circle_distance(pickup_lat,pickup_lng, pickup_1_lat, pickup_1_lng)*0.621371) for distance calc for both distance to each others pickups and drop-offs.

and date_diff('minute', created_time, created_time_1) as order_creation_delta

So something like any 2 orders or pairs of orders that are within 3 mins created of one another and 3 miles from one anothers pickup location and 3 miles from one anothers drop-off location. 

    with data as 
( select
    a.business_day,
        a.delivery_uuid,
        a.order_created_time_utc,
        a.pickup_lat,
        a.pickup_lng,
        a.dropoff_lat,
        a.dropoff_lng
from integrated_delivery.managed_delivery_fact a
where a.business_day between (timestamp '2019-03-01') and (timestamp '2019-03-03')
    union
    select b.business_day as b_business_day,
        b.delivery_uuid as b_delivery_uuid,
        b.order_created_time_utc as b_order_created_time_utc,
        b.pickup_lat as b_pickup_lat,
        b.pickup_lng as b_pickup_lng,
        b.dropoff_lat as b_dropoff_lat, 
        b.dropoff_lng as b_dropoff_lng
from integrated_delivery.managed_delivery_fact b
where b.business_day between (timestamp '2019-03-01') and (timestamp '2019-03-03')
)

stats as 
( select abs(date_diff('minute', a.order_created_time_utc, b.order_created_time_utc)) as order_creation_difference,
         (great_circle_distance(a.pickup_lat, a.pickup_lng, b.pickup_lat, b.pickup_lng)*0.621371) as pickup_distance,
         (great_circle_distance(a.dropoff_lat, a.dropoff_lng, b.dropoff_lat, b.dropoff_lng)*0.621371) as dropoff_distance
from data
)
select a.delivery_uuid, b.delivery_uuid, order_creation_difference, pickup_distance, dropoff_distance
    from data a
        cross join data b  
        WHERE a.delivery_uuid <> b.delivery_uuid
        and order_creation_difference <3
        and pickup_distance < 3
        and dropoff_distance <3

I have a query like the above, but not sure if I can calculate the values as a cte if I am unioning the tables prior?

Upvotes: 0

Views: 629

Answers (1)

Nicolas Busca
Nicolas Busca

Reputation: 1305

it seems that you need to do a join rather than a union.

with a as (select * from your_table)
select * from your_table
inner join a on 
great_circle(a.lat, a.long, your_table.lat, your_table.long) < max_dist
and abs(date_diff('min', a. date, your_table. date)) < max_time

explanation: an inner join of two tables outputs all and only combinations of rows for which the condition after on is true. You might want to impose a minimum distance as well to exclude matches of a row with itself.

Upvotes: 1

Related Questions