Area Normalize EPR Spectra

Hello,

I am relatively new to MATLAB and I am trying to area normalize these 25 spectra. By area normalize, I mean that the absolute value of the area under each individual spectrum is equal to 1. I have seen the previous responses to similar questions and I have attempted to replicate them with my data set.

The issue I am having is that although the original data (before area normalizing) all have similar peak to peak amplitudes, the area normalized data results in extremely variable spectral amplitudes. I realize that the amplitudes of different parts of the spectrum can change when area normalizing, but this kind of variability in the result does not seem plausible.

I included my matlab workspace so that the same data can be used. I also included a figure of the "incorrectly area normalized data".

Any help is greatly appreciated. Thank you in advance!

clc; clear; clf;

 Folder = uigetdir; %asks user for folder

Error using matlab.internal.lang.capability.Capability.require
Support for Java user interfaces is required, which is not available on this platform.

Error in uigetdir (line 52)
Capability.require(Capability.Swing);

 FilePattern = fullfile(Folder,'*.xml'); %searches folder for .xml files
 Files = dir(FilePattern); %puts files into struct array 
 SortedFiles = natsortfiles(Files);
 
     for K = 1:length(SortedFiles)
         thisfilename = fullfile(Folder, SortedFiles(K).name); %file name 
        
         [X,Y] = eprload(thisfilename); % extracts data
         
              
         DataY{K} = num2cell(Y);
         TablesY = cell2table(DataY{K}); % puts data into a table 
         ArrayY = table2array(TablesY);
         DataY{K} = real(ArrayY);         % takes only the real part of the array

         DataX{K} = num2cell(X);
         TablesX = cell2table(DataX{K}); % puts data into a table 
         ArrayX = table2array(TablesX);
         DataX{K} = real(ArrayX);         % takes only the real part of the array

         
         AreaUnderCurve{K} = abs(trapz(DataY{K}));         % absolute value of the total area under the curve 
         % (I took the absolute value so that it doesn't invert spectra)
         AreaNormalized{K} = DataY{K}./AreaUnderCurve{K};    % Divide each Y value by the total area under the curve
         
         AreaCheck = trapz(AreaNormalized{K});
         %plot(DataX{K},DataY{K})
         plot(DataX{K},AreaNormalized{K})
         legend

         hold on
     end 
     

[Image Analyst]

Can't see it here. Please post a PNG screenshot.

[Alex Garces]

Sorry about that. Here are those PNG files. Please let me know if there is anything else.

Thanks!

[Alex Garces]

This worked perfectly! Thank you for the help, it is very much appreciated.

I could be wrong but I don't think I can accept a comment as an answer. If you copy that to an answer I would be more than happy to accept it.

[Mathieu NOE]

hello

glad this helped

as you suggested, I moved it in the answer section.

thanks

Answer 1

[Mathieu NOE]

hello

I presume that negative signal area should aslo be counted as positive area

so this line is wrong

         AreaUnderCurve{K} = abs(trapz(DataY{K}));         

Should be :

AreaUnderCurve{K} = trapz(abs(DataY{K}));

otherwise a perfectly symmetrical sine wave of one period would have area = 0

Also I assume your x spacing is not 1 , then the area computation must also take that in account

area = trapz(X,Y) integrates Y with respect to the coordinates or scalar spacing specified by X.

         AreaUnderCurve{K} = trapz(DataX{K},abs(DataY{K}));