Filename to categorical in table

I created a datastore with accelerometer data of different files (S1_walking.csv, S1_sitting.csv and S1_laying.csv). Then I used to readall function to create a table with all data together.

Now, I want to add an categorical column containing a specific part of the file names to the corresponding part in the table. So next to every row of the 'walking data' a categorical 'walking' should be added, next to every row of 'sitting data' a categorical 'sitting' and next to every row of 'laying data' categorical 'laying'.

How can I do this easily by using the part of the filenames?

[Mathieu NOE]

I guess you want t make a string manipulation like this :

this will return 1 , so you can use it to fulfill (tick) the condition

filename = 'S1_walking.csv'
walking_tick =  ~isempty(findstr(filename,'walking'))

Answer 1

[Seth Furman]

Assuming we know which rows contain data from each file, we can create a categorical variable and add it directly to the table.

For example,

>> % Here we assume 'data' is our table.
>> % Rows 1-3 are 'walking data', 4-8 are 'sitting data', and 9-16 are 'laying data'.
>> data = table((1:16)')

data =

  16×1 table

    Var1
    ____

      1 
      2 
      3 
      4 
      5 
      6 
      7 
      8 
      9 
     10 
     11 
     12 
     13 
     14 
     15 
     16 

>> files = ["S1_walking.csv" "S1_sitting.csv" "S1_laying.csv"];
>> data.SourceFile = categorical([repmat(1,3,1);repmat(2,5,1);repmat(3,8,1)],[1 2 3],files)

data =

  16×2 table

    Var1      SourceFile  
    ____    ______________

      1     S1_walking.csv
      2     S1_walking.csv
      3     S1_walking.csv
      4     S1_sitting.csv
      5     S1_sitting.csv
      6     S1_sitting.csv
      7     S1_sitting.csv
      8     S1_sitting.csv
      9     S1_laying.csv 
     10     S1_laying.csv 
     11     S1_laying.csv 
     12     S1_laying.csv 
     13     S1_laying.csv 
     14     S1_laying.csv 
     15     S1_laying.csv 
     16     S1_laying.csv 

Alternatively, we could add the categorical filename data by calling transform on the datastore with a custom transform function.

https://www.mathworks.com/help/matlab/ref/matlab.io.datastore.transform.html

ds = datastore("sampleDir");
ds = transform(ds,@transformFcn,"IncludeInfo",true);
data = readall(ds)

function [data,dsInfo] = transformFcn(data,dsInfo)
    [~,filename] = fileparts(dsInfo.Filename);
    data.SourceFile = categorical(repmat(string(filename),height(data),1));
end

In my case this outputs the following

data =

  16×2 table

    Var1    SourceFile
    ____    __________

      5     S1_laying 
      6     S1_laying 
      7     S1_laying 
      8     S1_laying 
      9     S1_laying 
     10     S1_laying 
      5     S1_sitting
      6     S1_sitting
      7     S1_sitting
      8     S1_sitting
      9     S1_sitting
     10     S1_sitting
      1     S1_walking
      2     S1_walking
      3     S1_walking
      4     S1_walking