Column Header is a Date, but the content of Column is a value - How to reflect this in a histogram

[Anonymous]

Hi Folks, 

Wondering if someone can help.  I have an export of data from Primavera P6 in excel format.

Six of the columns contain text data, and the remaining columns headers are dates with the data for each entered.  See below screen shot.

I want to create a histogram, along the bottom axis should be the date, then stacked should be the Department with the a count of all data within that date against the department.

I am struggling to create this and wondered if someone could help point me in the right direction to where I am going wrong - the issue appears to be the date column being the headers and numerical values being held within the date (for reference the data is man hours held under each date)

[mauriciosotero · Accepted solution]

Hi,

In fact, the best practice is to transform theses Date columns in two: Date and Value (or a better name).

For that, firs you have to unpivot the date columns:

1) Go to query editor;

2) Select the date columns;

3) Choose the options Transform--> Unpivot Columns --> Unpivot the selected columns:

The result will be:

After, you have to modify the type of the new column date (in my example the "Atribute" column) to Date.

And that it! Chose and apply the changes and you will be able to do your visuals.

If this helps, I apreciate you give a kudo and mark as the solution. Thanks

[Anonymous]

In the event that the dates change in another dataset, this query will break. Is there a workaround for that?

[Anonymous]

Thank you so much! it worked.

[mauriciosotero]

Welcome!