Duplicate values on time span

[MaryDay]

Hello! I have a selection of customer names and dates of visits. In both columns the data can be repeated. I need to identify those customers who visited cafe more than one time in 14 days. I can't figure out how to do this. Please, help. Example:

11.06.2012    Mark  +

15.06.2012    John   -

17.06.2012   Mark   +

17.06.2012   Sam    +

29.06.2012   Sam    +

05.06.2012   Piter    -

05.06.2012   John    -

[Anonymous]

What does it mean "more than 1 time in 14 days"? Which 14 days? Where is the beginning and where is the end of the 14 days?

Best

Darek

[MaryDay]

@Anonymous  maybe you have some ideas? please 

[Anonymous · Accepted solution]

I think this is what you need:

Find the names of the customers that have 2 different visit dates not more than 14 days apart.

Let's say your table that stores visits is V, you've got CustomerID and VisitDate in there. Then you also have a dimension table with your customers C where you store unique CustomerID's. C joins to T in a 1:many fashion on CustomerID. Then you could add a column [2 Visits Within 14 Days] to C:

 [2 Visits Within 14 Days] = -- calculated column without the use of context transition
var __custId = C[CustomerID]
var __visitDates =
    SUMMARIZE(
        FILTER (
            V,
            V[CustomerId] = __custId
        ),
        V[VisitDate]
    )
var __2VisitsExist =
    NOT ISEMPTY(
        FILTER(
            CROSSJOIN(
                SELECTCOLUMNS (
                    __visitDates,
                    "FirstVD", V[VisitDate]
                ),
                SELECTCOLUMNS(
                    __visitDates,
                    "SecondVD", V[VisitDate]
                )
            ),
            [SecondVD] - [FirstVD] < 14
            && [SecondVD] > [FirstVD]
        )
    )
return
    __2VisitsExist

Best

Darek

[MaryDay]

Thank you very much for the answer. But the fact is that there can be more than two visits in 14 days and even several in one day, that is, the dates will be repeated.

[Anonymous]

-- calculated column without the use of context transition
[At Least 2 Visits Within 14 Days] =
var __custId = C[CustomerID]
var __visitDatesWithCounts =
ADDCOLUMNS(
    SUMMARIZE(
        FILTER (
            V,
            V[CustomerId] = __custId
        ),
        V[VisitDate]
    ),
    "CountOfSameDayVisits",
    	var __visitDate = V[VisitDate]
    	return
	    	COUNTROWS(
		        FILTER (
		            V,
		            V[CustomerId] = __custId
		            && V[VisitDate] = __visitDate
		        )    		
	    	)
)
var __2VisitsOnSameDayExist =
	MAXX(
		__visitDatesWithCounts,
		[CountOfSameDayVisits]
	) > 1
var __2VisitsOnDiffDaysExist =
    NOT ISEMPTY(
        FILTER(
            CROSSJOIN(
                SELECTCOLUMNS (
                    __visitDatesWithCounts,
                    "FirstVD", V[VisitDate]
                ),
                SELECTCOLUMNS(
                    __visitDates,
                    "SecondVD", V[VisitDate]
                )
            ),
            [SecondVD] - [FirstVD] < 14
            && [SecondVD] > [FirstVD]
        )
    )
return
	__2VisitsOnSameDayExist || __2VisitsOnDiffDaysExist

Best

Darek

[Anonymous]

My code does not tell you that there were 2 visits within 14 days but that there were AT LEAST 2 different dates no more than 14 days apart on which the person visited. This is how I understood your description.

If you need something different, then you could try to adjust the code. From what you've replied it looks like you don't want to check if there were at least 2 different days within a 14-day span but that there were at least 2 visits regardless of whether on the same day or not.

Can you please confirm?

Best

Darek

[MaryDay]

It should be a cycle that checks every 14 days from the first date. That is, from 01.06 to 14.06, then from 02.06 to 15.06, and so on. I need a list of customers who visited cafe more than once in 14 days at any time.