[Don't merge] Octave-compatible changes

[ghost]

Hello James,

as per your request, here are the changes I made to make your ImportAgilent.m code compatible with Octave 4.0.1 (as it is shipped with Debian 9.2). I thought that a git diff would be a much better way to show the changes, that's why I made a pull request. I don't expect you to merge these changes, especially since they break the inputParser, but it is much more comprehensible that way.

I also added comments explaining the reasons why I had to made that particular change, but for simplicity's sake, here is a summary:

• addParameter (method of inputParser) does not exist in Octave (Version 4.0.1). Here is the corresponding issue. It looks like they have some patches available, but I was not sure about that. Also, I did not need the functionality and it was easier for me to just short-circuit it. I don't know if there's an alternative for that
• In Octave, you use javaObject('fully.qualified.class.name', args...) to initialise a Java object. It seems like this functionality is also available in MATLAB - at least I found a source/documentation where they explicitly mention MATLAB
• The Java package com.mathworks.* obviously doesn't exist in Octave, but the functionality you want is implemented in the native Java libraries, namely Swing. Initially I just dropped the filter code, because I just picked the .CH files and that's it, but I really think that this is more elegant since it uses native libraries instead of some MATLAB façade.

Now that I am writing this PR I see that I didn't properly change line 349:

fc = javax.swing.JFileChooser(java.io.File(pwd));

To exactly replicate the behaviour, you'd need to do it like that: (may not work, that's from the top of my head)

fcParam = javaObject('java.io.File', pwd);
fc = javaObject('javax.swing.JFileChooser', fcParam);

I am aware of the fact that the javaObject syntax is not as pretty as writing the fully qualified class name like it would be Java-native, but at least it works in Octave as well.

Please let me know if there's anything I can help you with.

Greetings,
Chris

[ghost]

I now added the Python plot script and an example CSV data file. The data was generated with your code, namely ImportAgilent.m and ExportCSV.m.

data = ImportAgilent()
% Then I chose the .CH file in the file choser
ExportCSV(data.time, data.intensity, 'file', 'hplc-out.csv')
% Please note that I used data.time and data.intensity, regarding your last message. 
% So it should already be corrected (corrected_signal = raw_signal * slope + intercept).
% Data.time was returned in [min], but as you later corrected me, it is expected to be
% in [s] for the peak area calculation. Therefore there's an `* 60` in the Python plot script

However, the file name last parameter did not work correctly. the file was always named data.CSV. I am not entirely sure, but I think I accidentally downloaded ExportCSV.m from master, while ImportAgilent.m was from dev. However, it worked, I just renamed the file and that's it. Please let me know if you also want to take a look at the original .CH files - I cannot add them publicly because they contain personal information, but I can supply them via email.

Greetings,
Chris

PS.: To execute the Python script, use a Terminal/CMD (depending on your OS) and go into the directory where the .CSV example file is located. Then run python plot.py. You need to have Python, and the Python-libraries installed for it to work, namely matplotlib and scipy. (I used Python 3.5, but Python 2.7 should work, too, I think...)

[ghost]

If you're interested, here are the locations where I used the information you gave me in your last message.

[ghost]

This is where I subtract the baseline from each y-value

[ghost]

Here I search for the minimum y-value for the baseline

[ghost]

Here I convert the minutes to seconds

[ghost]

Here I create a subset of the original data, that is, the peak itself limited by PEAK_START_TIME and PEAK_END_TIME. As you already said: This is subjective, but it is precise enough. The constants are defined at the beginning of the script. Please note that the values will be refined later, it was just an approximation to validate the calculation.