Last Year Sales Comparison Based On The Given Date

[dineshm030]

Hi Experts,

I have created the Current Year's and Last Year's sales reports separately. I need to show the last year's sales based on the given date.

I am maintaining the separate source file which is given below.

Working Day	CY DateKey	LY DateKey
1	20230301	20220301
2	20230302	20220302
3	20230303	20220303
4	20230304	20220304
5	20230305	20220304
6	20230306	20220305
7	20230307	20220307
8	20230308	20220308
9	20230309	20220309
10	20230310	20220310
11	20230311	20220310
12	20230312	20220311
13	20230313	20220311
14	20230314	20220312
15	20230315	20220314
16	20230316	20220315
17	20230317	20220316
18	20230318	20220317
19	20230319	20220318
20	20230320	20220318
21	20230321	20220319
22	20230322	20220321
23	20230323	20220323
24	20230324	20220324
25	20230325	20220325
26	20230326	20220325
27	20230327	20220326
28	20230328	20220328
29	20230329	20220329
30	20230330	20220330
31	20230331	20220331

If we are in 20230310, we need to show  till 20220310 sales whereas we are in 20230313, we need to show till 20220311 sales. We need to show last year sales based on the given date table.

Sum({<Year = {$(=max(Year)-1)},Date_Key = {"<=$(=Date(AddYears(Date#(Max(Date_Key),'YYYYMMDD'), -1), 'YYYYMMDD'))"}>}Sales_Value)

I have Date table in my data model and sales table has a date key column.

@sunny_talwar @MayilVahanan @marcus_sommer 

Thanks in advance.

[marcus_sommer]

In general yes - but not with the "classical" association of data else you would need some extra logic, for example by using The As-Of Table - Qlik Community - 1466130. I doubt that's really expedient for your task else rather an unneeded detour - I think the extra condition of: Date_Key=p([LY DateKey]) is more causing problems instead of being helpful. 

[dineshm030]

Could you please provide the expression

[marcus_sommer]

Maybe:

Sum({<Version_Desc = {'Actual','Adjustment'},ContinuousDay = {"<=$(=max(ContinuousDay))"}, YearOffset = {"$(=max(YearOffset )-1)"},Year,Month,Day_of_Week,Date_Key>}Sales_Value)/1000

[dineshm030]

Hi,

When I am using this expression,  that highlighted rows have to exclude when we select the 20230314

[marcus_sommer]

Like above hinted I suggest not to use the date_key as dimension else a dimension without the year-information. If there are further differences between last and current year it would mean that the match on the working-day isn't sufficient.

[dineshm030]

Even the text box is showing with 20220313 and 20220314 sales value.

[marcus_sommer]

I think it means that the created ContinuousDay didn't match your expectation. Just put the calendar-fields:

year, month, week, day, date, DD-MM, DayNumberofYear, ContinuousDay, YearOffset

in a table-box and filter various ContinuousDay values directly and/or in combination with other period-fields.

If you select 20230314 and afterwards the possible value of the ContinuousDay and de-selecting 20230314 again you should see at least two rows - at least one from the current year and at least one from the previous years - and all shown dates needs to fit to your expectation. If not you need to adjust the creating-logic within the calendar until it worked for all dates.