New feature requests - General

[Robbert van de Kruijs]

David,

 

I guess we should not start a new thread for each feature request, so I'll start with a "general" topic.

Below is a few things I encountered while using the new IMD, which led me to request "new features", hope they can be easily implemented :)

 

- In the old IMD (4.1.1), there was a file "imdsave.dat". This has apparently been replaced by "default.imd" (or an auto-generated name if that option is on), but this new file does not auto-load on start of IMD like the old one: bug or feature? I really liked the old way, did I miss something to enable this in the new version?

 

- Preferences do not auto-load either on new start of IMD, you have to manually load the ".ipr" file (or load an "*.imd" file where they appear to be included. Is that correct? Why not autoload a default configuration?

 

- In preferences, "Automatically save data" is standard disabled, why cannot this option also be saved in preference file?

 

- In IMDXPLOT, additional controls are standard hidden, would be nice to have option in preferences to have it standard hidded or standard visible. (previous IMD had things like log scaling right next to graph, I miss that option, don't want to click to get it back :)

 

- Would be nice to be able to plot(and save) a "delta" profile, I did not find it (but am still discovering new things each day, might even have to read the manual ...), and in this profile it would be really nice to include gradual change at interfaces due to roughness.

 

Cheers,

Robbert

 

[David Windt]

Hi Robbert,

I'll respond to each of your feature ideas in turn. Please see below.

David

On Monday, October 28, 2013 4:58:34 AM UTC-4, Robbert van de Kruijs wrote:

- In the old IMD (4.1.1), there was a file "imdsave.dat". This has apparently been replaced by "default.imd" (or an auto-generated name if that option is on), but this new file does not auto-load on start of IMD like the old one: bug or feature? I really liked the old way, did I miss something to enable this in the new version?

The auto-load capability in the old version is actually no longer included. It's something that I personally didn't use any longer, and so I just got rid of it, not knowing how many other people used it. I would be happy to consider putting it back in, but before I do, I'd like to get a better sense about how many people really want it.  So: IMD community, please speak up about this -  Do you want it back or not? (Before you answer, please keep in mind that you can auto-load IMD Preferences at start-up, as I'll explain next.)

 

- Preferences do not auto-load either on new start of IMD, you have to manually load the ".ipr" file (or load an "*.imd" file where they appear to be included. Is that correct? Why not autoload a default configuration?

You can indeed auto-load a .ipr (IMD Preferences) file at start-up. How to do that is explained in section 10.3 of IMD.pdf.
 

 

- In preferences, "Automatically save data" is standard disabled, why cannot this option also be saved in preference file?

I decided to make the three File Saving preferences 'persistent': when you open a .imd file, those settings do not get changed. However I didn't give any thought to these settings with regard to opening a .ipr file. But now that you raise the question, I agree that it's a good idea to have these settings included in .ipr files so that they will be restored when importing preferences - thank you for that!  I'll include that in the next version.
 

 

- In IMDXPLOT, additional controls are standard hidden, would be nice to have option in preferences to have it standard hidded or standard visible. (previous IMD had things like log scaling right next to graph, I miss that option, don't want to click to get it back :)

This is also a good idea - I will do it for the next version!

 

 

- Would be nice to be able to plot(and save) a "delta" profile, I did not find it (but am still discovering new things each day, might even have to read the manual ...), and in this profile it would be really nice to include gradual change at interfaces due to roughness.

I don't understand what you mean here - can you please explain further?
 

 

Cheers,

Robbert

 

[Robbert van de Kruijs]

Hi, David,

 

With a "delta" profile, I mean to plot the in-depth distrubution of (1-n) (n=optical constant :)

Normally this is a "step function", but with roughness, the steps get smoothened, you can take an errorfunction to "connect" the bulk values. This gives a good idea on what your x-rays are sensitive for, could for example immediately see that there's a small step going from material X in one layer to material Y in next layer, something you don't see from density difference alone (e.g. different density could be compensated by difference in atomic scatterinf factors). The addition of roughness in the profile also gives visual confirmation if the rougheness is actually large compared to layer thicknesses (and could/should serve as a warning sign that traditional debeye waller / nevot-croce / whatever mathemetical representation of roughness starts to break down... During my PhD I actually always took the delta profile and as a "reality check" disables beta (no absorption) and plotted R+T, which clearly starts to deviate from unity if roughness it too large compared to layer thickness. A visual check would be simple and fast :)  GenX also nicely plots this "scatter length density" profile for delta and beta.

... I'm being silly here, a picture says more than a thousand words, so here it is : http://tinypic.com/r/sx2es5/5

 

 

 

 

And talking about feature requests: In GenX, apparetly Matts is able to estimate some parameter confidence interval during the fitting procedure. This helps tremendously in estimating which parameters still have large uncertainlties, and therefore tells you when you can stop the optimization routine. Picture: http://tinypic.com/r/303jr61/5

 

Cheers,

Robbert

 

[Markus Kraemer]

The auto-load capability in the old version is actually no longer included. It's something that I personally didn't use any longer, and so I just got rid of it, not knowing how many other people used it. I would be happy to consider putting it back in, but before I do, I'd like to get a better sense about how many people really want it.  So: IMD community, please speak up about this -  Do you want it back or not?

I would like to have it back :-) and my colleagues as well. It was quite convenient.

 

 

- Preferences do not auto-load either on new start of IMD, you have to manually load the ".ipr" file (or load an "*.imd" file where they appear to be included. Is that correct? Why not autoload a default configuration?

You can indeed auto-load a .ipr (IMD Preferences) file at start-up. How to do that is explained in section 10.3 of IMD.pdf.
 

Unfortunately that only works if you run IMD from IDL and not if you just use it as an XOP extension (like me). That's why the old option was quite useful.

[Dmitry Kuznetsov]

Hello, David.

A small suggestion - it would be convenient for me to see the path (through OS folders) and the file name
which is currently opened in IMD somewhere on the main window.
Because if you fit a lot, and try different parameter intervals or smth. for the same
data (and, for instance, different sets of data for the same experiment - like reflectance peaks for different multilayer periods), sometimes a lot of additional efforts are needed not to mix up.

Probably, it has sense.

Cheers, Dmitry.

[Dmitry Kuznetsov]

Also, a design suggestion: when there are a lot of fitting parameters,
and one wants to look through their changes during the fitting, he should manually fit
the columns width to see all the columns at the same time.
Picture tells better (upper part - initial view, lower part - how it is convenient):
https://drive.google.com/file/d/0B6T92xyewYcpMGtSdzJfTS1iMGc/edit?usp=sharing
Maybe, that's useful to have automatic fit of column width.

Cheers, Dmitry.