1 of 1 people found this helpful
Take a copy of order date from columns self and place it in the filters. Right click and select Add to context
Now the order date will turn into gray color on the filter
On dashboard right click ship date and select all values in context
On next filters ship date 2 and order date 2 select only relevant values
thanks for the quick response and your solution!
Unfortunately it does not yet deliever the correct results. The marked value for Oct17 in sheet1 should be the same as in sheet2.
What I forgotten to say, is that sheet2 is only for checking the values from sheet1. It simply counts the number of records filtered by ship date and order date which has to be adjusted manually to one time frame (for example Oct17 + prev. 11month).
As already one ship date is there then why you have added ship date 2?.
The order date might differ from ship date because shipping will be done after ordering item.
As of now i have fixed the max values of sheet 1 and sheet 2 are same. Let me know is this solve your issue
I'm sorry, the example I created obviously doesn't really help and I think I could not yet explain well enough what my goal is. I would like to try it one more time.
Maybe this would be a better headline: Moving sum with two different date dimensions?
My goal is to visualize the trend of the field failures (of products) in monthly values (and later the field failure rate). Each row of the data is one field failure. Each monthly value shall be the sum of the rows of the previous 30 month (why I'm using moving sum). Now the tricky part: I have to consider 2 different dates, the "failure date" and the "sale date" (in the example it would be ship date and order date). So only the field failures which have failure date AND sale date in this 30 month time slots shall be considered.
Is there a possibility in Tableau to do that?
I would know how to build the moving sum with one date dimension, but not with two. I would know how to get the value for one single month (simply two date filters), but not for the Trendline.