Dwell time analysis of network

[Anonymous]

Hi, I have data by transport transactions. A parcel can travel mutiple legs before it reach the end location. I want to make a dwell time analysis of the nodes in the network the parcels passes. For each location (A, B, C, etc) how long time are the parcel staying there in average?

Article	Leg start location	Start time	Leg End location	Arrival time
1	A	2020-01-01	B	2020-01-03
1	B	2020-01-05	C	2020-01-06
2	A	2020-01-01	B	2020-01-03
2	B	2020-01-04	C	2020-01-05
3	A	2020-01-01	B	2020-01-03
3	B	2020-01-04	D	2020-01-04
3	D	2020-01-05	C	2020-01-06

Any good idea?

[V-lianl-msft · Accepted solution]

Hi @Anonymous ,

First, I added index columns to the query editor to get the last arrival time.

Create measure to get the last arrival time and calculate the time difference.

Measure 2 =
VAR current_index =
    SELECTEDVALUE ( 'Table'[Index] )
VAR last_a_date =
    CALCULATE (
        FIRSTNONBLANK ( 'Table'[ Arrival time], 1 ),
        FILTER (
            ALLEXCEPT ( 'Table', 'Table'[Article] ),
            'Table'[Index] = current_index - 1
        )
    )
RETURN
    DATEDIFF ( last_a_date, MAX ( 'Table'[ Start time] ), DAY )

Calculate the average time of each location.

Measure 3 = AVERAGEX(ALLEXCEPT('Table','Table'[ Leg End location]),[Measure 2])

Sample .pbix

Best Regards,
Liang
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.