HELP - How to make a line chart that shows the improvement of certain occurrences over time?

[Anonymous]

Hey there!

I'm trying to make a line chart that show the improvment of occurrences (issues / problems) over time.

I have a table with the date when the issues were detected and when said issues were fixed (if they got fixed). The idea is to show the development of the issues trough time, with the line going up with new issues and down with the closing of older issues. Something like this:

The table also have other fields that can be used as filters, like the issues priority (critical, medium and high) and their status (open and closed). The most important filter here is the status, I think, because this is how I know the problem is still unsolved.

The problem his, I can only make the graphics with the issues associated with the date of creation or date of resolution, and they need to have a span in time, if the issue was open for 9 days, then the line should be up for 9 days with that issue. Instead what I have is the line going up on the date that is associated with the issues (ie. if I have 2 issues on March 3rd, the line goes up 2 values, if on the following days I have no issues the line come down to 0, even though the issue is still open). This is what I have:

Can anyone help me? Appreciated.

[vanessafvg]

this really has to do with the way your data is modelled.  If you want an issue to be persisted until its resolved, you would probably need a row for every day you are tracking  it as unresolved until it is resolved.  Your data point only says issue on the day 1.  It needs to know the full length of time that is is an issue.   

Doing something with a cummulative count per day,  to add and subtract new issues and resolved issues, or having a row for each day it is open would give you the number you are looking for. But it requires some data modelling.  There are probably other ways of doing it too but this is my suggestion.

You could probably also create a measure but you would need to provide some sample data to see if that would work.