Wish List for PFE Features

[John Donovan]

I tested both the Imaging at Acquire! and Digitize Image and both work nicely.
Thanks a lot.

Awesome!  Thanks for checking. 

We must be doing something right, because we only had to add three lines of code in the window "Resize" event, and it all just worked fine!   

[Ben Buse]

Hi John,

Can you remove the minimum peak to background of 1.5. I want to peak up on Fe Lb and Fe La, and I can't achieve 1.5 over the window for peak search. I'm currently doing wavescans first but this is not ideal

Thanks

Ben

[John Donovan]

Hi John,

Can you remove the minimum peak to background of 1.5. I want to peak up on Fe Lb and Fe La, and I can't achieve 1.5 over the window for peak search. I'm currently doing wavescans first but this is not ideal

Thanks

Ben

Hi Ben,
I'd prefer to not remove it completely. How low a P/B do you need do you think?  1.4? 1.3? 1.2?
john

[Ben Buse]

Hi John,

Sorry I've just looked at it and it should be 0.24/0.12 so 2 and greater, I've currently got the minimum background value set to 2, I think I'll change it to 1.5, and make sure the shoulder of the Fe La is not in the peak window when peak searching for Fe Lb, I'll see how this goes

Thanks

Ben

[Ben Buse]

Hi John, I can change the minimum background to peak to 1.5 or 1.6 in the software for the current run, but the scalers file will not accept it, it states invalid property. In scalers I currently have 2. which works

One separate thing for wish list - when exporting standard compositions - it be nice to export published values

Thanks

Ben

[Ben Buse]

Hi John,

I'm copying and pasting in excel, and I have several wishes for exporting data

(1) the suggestion above - exporting published quant.

In addition:

(2) allow export only of average composition and std dev.

(3) the most important - suggested by several users - format data from multiple setups - at the moment if the elements change the columns change and its interrupted by headings. Would it be possible to have the same columns for all the data exported - i.e. look through the list of samples to be exported and pick out the total number of elements - and use these as the columns, OR export the different setups to different files?

Sorry this sounds like a lot of work - I don't know how many people would find pts 1 and 2 useful.

Thanks,

Ben

[John Donovan]

Hi John, I can change the minimum background to peak to 1.5 or 1.6 in the software for the current run, but the scalers file will not accept it, it states invalid property. In scalers I currently have 2. which works

Hi Ben,
Yes, that is correct.  I'm trying to be conservative and in general prevent people from having a dangerously low minimum P/B for peaking in all runs.

However, as you say, having a specific run with a P/B as low as 1.5 can be specified by the user in the app.  And if you create a new run with that low P/B minimum value, it will get loaded into the new run automatically when you use the Load File Setup option.

Will this work for you?
john

[John Donovan]

Hi John,

I'm copying and pasting in excel, and I have several wishes for exporting data

(1) the suggestion above - exporting published quant.

In addition:

(2) allow export only of average composition and std dev.

(3) the most important - suggested by several users - format data from multiple setups - at the moment if the elements change the columns change and its interrupted by headings. Would it be possible to have the same columns for all the data exported - i.e. look through the list of samples to be exported and pick out the total number of elements - and use these as the columns, OR export the different setups to different files?

Sorry this sounds like a lot of work - I don't know how many people would find pts 1 and 2 useful.

Thanks,

Ben

Hi Ben,
I want to make sure I'm understanding exactly what you are requesting.

This copying and pasting into Excel- you're doing this from the Analyze! window using the little "Copy" buttons?
john

[Ben Buse]

Hi Ben,
Yes, that is correct.  I'm trying to be conservative and in general prevent people from having a dangerously low minimum P/B for peaking in all runs.

However, as you say, having a specific run with a P/B as low as 1.5 can be specified by the user in the app.  And if you create a new run with that low P/B minimum value, it will get loaded into the new run automatically when you use the Load File Setup option.

Will this work for you?
john

Yes that's fine thanks

Ben

[Ben Buse]

Hi Ben,
I want to make sure I'm understanding exactly what you are requesting.

This copying and pasting into Excel- you're doing this from the Analyze! window using the little "Copy" buttons?
john

Sorry I should have made it clear, I'm reformating the data once I've exported it using the user specified output script, and I was wondering if you could refine it?

Thanks

Ben

[John Donovan]

(1) the suggestion above - exporting published quant.

Hi Ben,
Thanks for the clarifications.  This option already exists in the user specified output:



The only caveat is that the software doesn't select automatically select the proper Standard/Unknowns output option when exporting data from the Analyze! window and using the right click option to export only selected samples.  The option is selected on the basis of what output you last did from the Output menu.

I just fixed the code to do this automatically and will upload a new version later today.  In the meantime just use the Output | Export User Specified Output menu for standard sample output with the published values options.

(2) allow export only of average composition and std dev.

Yeah, not a bad idea, but the problem is that the user has to select as output type so it knows what data to average. I'll have to think if there's a clever way to do this, but for now you'll have to remove the other rows in your spreadsheet.

(3) the most important - suggested by several users - format data from multiple setups - at the moment if the elements change the columns change and its interrupted by headings. Would it be possible to have the same columns for all the data exported - i.e. look through the list of samples to be exported and pick out the total number of elements - and use these as the columns, OR export the different setups to different files?

I can see that this would be useful, but it would be a lot of work. I'll have to think about it how it might be done.

In the meantime you're a super scripter!  I'm sure you could easily knock off an Excel macro to do this...   

Post it to the forum and you'll be a hero! 

[John Donovan]

Related to Ben's request above we've modified the code to automatically select the proper sample output type (standard vs. unknown) when the user right clicks the selected sample list in the Analyze! window for the "User Specified Output Format" as seen here:



This enables the output of "published" concentration (and variance and difference) values for standards if standard samples were selected.

[Ben Buse]

Related to Ben's request above we've modified the code to automatically select the proper sample output type (standard vs. unknown) when the user right clicks the selected sample list in the Analyze! window for the "User Specified Output Format" as seen here:



This enables the output of "published" concentration (and variance and difference) values for standards if standard samples were selected.

Thanks for doing that

Ben

[Probeman]

The "Email notification" box is also greyed out. I guess I have to put in the accurate SMTP etc notes in the probewin.ini file first which I haven't done yet because I thought it only reports errors  . The F1 Help file here also gives no additional information FYI.

All,
I am bringing this post "back from the dead" because Paul Carpenter recently asked me if this email report feature still works and I thought, well I haven't tested it in a while, so I'll do that.  And lo and behold it does still work- at least on our Univ of Oregon email server. 

For those who are not yet familiar with this feature, it is accessed from the Acquisition Options from the Acquire! window as seen here:



The Probe for EPMA help file has this to say:

SMTPServerAddress="smtp.probesoftware.com"
SMTPAddressFrom="donovan@probesoftware.com"
SMTPAddressTo="donovan@probesoftware.com"
SMTPUserName="donovan"

These keywords are to be used for e-mail notification of real-time automation errors. If these SMTP fields are valid and the option "E-mail Notification of Errors" is checked in the Automation Options dialog, then any real-time errors (for example, a blown filament message) will be trapped and forwarded to the address indicated. This option is only available for automation procedures initiated from the Automate! window.

          SMTPServerAddress: must be a secure email server
          SMTPAddressFrom:   only used for documentation purposes

In addition, a "progress report" message will be sent every 8 hours to the e-mail address specified in order to give positive notification of the normal functioning of the instrument during especially long runs.

Note that the first time the automated acquisition is started and the E-mail Notification of Errors option is checked then the program will prompt the user for the password for the specified user email account. This password will be retained only for the duration of the currently open probe run.

The idea being that when this box is checked the software will send email reports every 8 hours or so on the progress of any automation running in Probe for EPMA. It will also send any errors it detects such as the filament burning out or a vacuum problem.

To test this email reporting feature I entered the relevant information in the [general] section of the Probewin.ini file as seen here:

[general]
SMTPServerAddress="smtp.uoregon.edu"                    ; email server address for error messages
SMTPAddressFrom="epmalab@uoregon.edu"
SMTPAddressTo="epmalab@uoregon.edu"
SMTPUsername="epmalab"

Of course you'll need to use your own email server and addresses!

When you start an automation in Probe for EPMA (in this test I simply confirmed some unknown positions), the software prompts you once for the email password and sends a "Automation initiated" email, and when the automation finishes it sends an "Automation completed" email message.

And as previously mentioned, if the automation runs longer than 8 hours it will send "progress" reports as to which sample it is working on.  Here are the two email messages I received in the probe lab email account for the test I performed:

[Paul Carpenter]

John,
Thanks for posting on this. The capability is good. Keeping the password only for the current session is not convenient. We use an email account for the microprobe at WUSTL and I would favor having some way to store the password but I can see the general IT-freakout of such a policy. Is it possible to securely push an email to an address like you have set up in the ini file?

The status report would be great. It seems to me that PFE is so bullet-proof that actual error messages for an automated run are hard to envisage (since the points are scanned before starting the run). Excessive drift in automated re-standardizations would be good to trap.

I would like to know when the run is finished because we don't use the wav file for this purpose; email is infinitely better for this purpose.

Cheers,

Paul