Problem to fit Y position annotation label Offset, help!

[Fonsi]

Hello,

I am using histogram to draw several values and I want to use annotations to show the value of each bar.

I fit the Y scale to have space for the annotation value using the max value of data. But I can't fit the label offset to have same free space in all the bars. If you see the attached file, you can see the numbers cover the bar and they dont look ok. Do you know any way to fit to get all the values annotations with the same free space with the bar?.

Thanks for all.

Fred.

 

[ben64]

Set the label offset to 0,1 for all, it will do the trick.

edit: Set the y offset value to a smaller number to have the annotation closer to the top of the bar.

Ben64

[Fonsi]

Hi Ben,

Thanks for reply!.

Yes, I can to play with the closer value, but It is only one example. I use 7 histograms and its part of one sub-vi the values will vary. I need any way to programmatically fit the y offset value. It seems no sense.

Sometimes with 2 is ok all the values, if its > 10 i need 3... i dont understand.

[ben64]

I don't understand the issue, the offset is relative to the top of the bar. If the bar height change the distance between the top of the bar and the label stays the same. Why does it sometimes need to be different?

Ben64

[Fonsi]

Ben,

No, You can check in the example vi. Change the input values and you will see that with diferent values the gap between label and bar is not the same.

Upper example, the first value is 15, and the offset label is 2. You can see is no look good. In the bottom example, with value 9, the same offset label and show good.

[altenbach]

There are many things wrong with your code and you are doing this way too complicated.

• To add an offset to the label, you would use addition, not multiplication.
• A plain waveform graph is sufficient, since the x values are equally spaced. No need for an xy graph.
• If you use a spaced bar plot style, you don't need to interlace zeroes.
• If you would normalize the data to [0..1] Things get easier. You can still annotate with the original values.
• You only need to bundle the values that differ. All others can be set in the cluster diagram constant once.

Here's a quick example. Modify as needed.

 

[altenbach · Accepted solution]

Of course if certain elements are constant (e.g. the number of bars) many properties need to be written only once before the loop. Here's the code in action using generated random data.

 

[Fonsi]

thank you so much, altenbach!.

You are the man, it works very good. You are right my design was very complicate.

Yes I use a fixed value of 10 bar. I put  interlace zeroes to do the bars thinner. So the number of annotation is more clear. But your code is better and easier.

Thank a lot!.

Fred

[altenbach]

---

Fonsi wrote:

Yes I use a fixed value of 10 bar. 

---

Then just make the y scale from 0..14 and add 2 for the offset. No need to divide by the max. (might need to fine tune based on the exact font size).

[Fonsi]

Ok, thanks for help!.

Fred