Is there a question in there somewhere?
Is it possible to figure out this need?
It even carries a question mark at the end
Perhaps it is my english fault. I should have said solve this out instead of figure this out.
Test Test, first might I suggest you use your actual name. This gives us a hint that English may not be your primary language, and you get a whole lot more leeway from us. I did see the question (even the question mark ) but I'm not sure what that question means. What result are you getting? What is not working the way you want it to?
Here's where the extra 'leeway' comes in. I'm going to guess (having taken a closer look at your measure) that you are trying to get blank months to show as zeros. If that's the case then try using:
That should do the trick.
I tried to change my name, but I could not in my menu. I'll try again.
Thanks for your answer, ZN is the same as my IFNULL I guess. But the problem does not come here but how to do the whole picture.
I mean if I drag Date as continous, it shows a visual axis. If I do it as discrete, only the rows with value are shown.
My problem is getting my date over a discrete (because it is composed by individual values), but continous (because it shows any possible value and not only those present actually) Date Axis
I'm thinking that you want to show 0s for the missing months? In that case, on your discrete date pill, turn on "Show Missing Values" to pad out the data. Then on your measure pill, click on that to go to Format->Pane Tab->Special Values section->Text box and set that to 0.
Thanks Jonathan but two problems rise:
a) I can only show missing values between two most extreme data entries. If I have Jan,Feb,June for NY Revenues and Feb, May for Washington Revenues, they won't show uniformly all values from Jan to June. NY will now show March, April and May, but Wahington will add March and April but no January and June, as they are outer to the inner data range, not a cartesian product among all domains.
b) ZN works well if the value is null, but if the whole register is missing, the default value is not shown
This is what you said you wanted in your original posts:
If that is what you actually want, then see attached. If that is NOT what you want, post a packaged workbook with sample data that demonstrates your actual situation, with a better explanation (screenshot) of what you actually want. Thanks,
test.twbx.zip 13.9 KB
There may be ways to get what you want using a combination of domain padding and/or domain completion, however they depend on pill arrangement and the desired view. Like Shawn said, we'd need to see a packaged workbook with some sample data that matches your scenario and a better outline of your desired results.
Hi Ramon, nice to see you on the forums! Scaffolding is in the back of my mind as a potential solution as well. I hadn't mentioned it because that kind of scaffolding can be a lot more work to set up, particularly since the view needs to be partitioned/filtered in some not-yet-clear way.
Hi Jonathan, I'm just starting to do some contribution on the forums.
I agree with you regarding the use of scaffolding technique. We need to explore simpler solutions.
Thanks for the follow.
I have no workbook because I am only using this for learning.
I could lose the time doing some sample data and book, but again, we are in the same situation, I cannot show what I don't know how to do. Indeed I could excatly pass your very same file.
I need to have a fixed domain range COMMON to a discrete dates, like if they were continous
In the example you passed me, date entries where defined for each month. What if you did not have data for December but even though you want to show it because you want to show all months for that year (= continous range) in a discrete way?
Show missing values let you show any intermediate month value missing. But I want to define the range statically, and show all the missing data as zero. Even if the record for that city start at June, I want this and all cities show discretly in a table from Jan to Dec (zero if no data, not only NULL value, but MISSING as well)
I am just wondering if there is a way to acomplish it. No need for a concrete solution, just the technique/approach in words to do my own research.
Thank you and sorry!
And you thought we'd get all this from your original post?! I'm done.
Thanks Jonathan. Domain completion is the word for the technique I am looking for probably. Let's research it a bit. If you have any suggestion, I am open!
You are right Shawn. What is confusing in my original post is that I filled data until June. If I had put July as well in the example, then you could have guessed it was not a matter alone of "showing missing values"
I thought the title was pretty much self-explanatory:
Discrete Date over "discretized continuous date axis"
I have a discrete Date and I want it over a discretized continous date axis.
If you have a continous date axis, then it has a visual range like "all months for 2013".
If you discretize a continous date axis, it will be something like Jan,Feb,March,Apr,May,Jun,Jul,Aug,Sep,Oct,Nov,Des
If you put your discrete date over (which does not feature all values looking my sample), you would have Jan, Apr, May, but the rest does not show as NULLS or Zero, at least in the extremes and after "showing missing values".