Compare Previous Day Data to Today's Data

[gmasta1129]

Hello,

I have a simple table which pulls data in each day (except the weekends).  

The table consists of the following columns 

1. Customer Number

2. Time Stamp

3. Contract Number

I am looking for a formula to compare  today's contract number to previous days contract number and if the contract number is found then pull in "OK". If the contract number is not found then pull in "Missing".

For example, in the table below, the contract number "000FXFW230270024" can be found on both 2/6 (today) and 2/3 (previous day) files, therefore, the column should state "OK" but if it was not found on 2/3 file, then it should say "MissIng".

Note: This should always be compared to the previous day.  For example,

1. 2/6 data should compare to 2/3 data

2. 2/3 data should be compared to 2/2 data.

3. 2/2 data should be compared to 2/1 data

cust #       time stamp      contract number

55555          2/3/2023          0000FXFW230270024

55555          2/6/2023          0000FXFW230270024

44444          2/3/2023          0000FXFW230270023

44444          2/6/2023          0000FXFW230270023

[smpa01 · Accepted solution]

@gmasta1129  sure

Measure = 
VAR prev =
    CALCULATE (
        [__contract],
        OFFSET (
            -1,
            DISTINCT ( ALL ( 'Table' ) ),
            ORDERBY ( 'Table'[timeStamp], ASC ),
            KEEP,
            PARTITIONBY ( 'Table'[cust] )
        )
    )
RETURN
    SWITCH (
        TRUE (),
        ISBLANK ( prev ) = TRUE (), "OK",
        [__contract] = prev, "OK",
        "Missing"
    )

[gmasta1129]

@smpa01 , thanks for the quick response, the formula has been trying to calculate for the past 10 minutes.  The column is blank.  Not sure what the issue is but please note, that there can be more than one contract number for each customer.  for example, customer number 55555 can have 10 different contract numbers in one day. (2/6/2023) Not usre if the max in the _contract formula is causing the issue. 

[smpa01]

that there can be more than one contract number for each customer--- the sample data is not representative of this