How to place NaN instead of " " / [ ] or empty field by using standardizeMissing?

Hi, everyone

Actually i already ask about this topic at : https://www.mathworks.com/matlabcentral/answers/1455464-re-write-a-data-from-file-readtable-readmatrix-etc-with-a-replacing-rule-for-a-specific-data-num?s_tid=srchtitle

I have tried standardizeMissing by using Readtable at that link (I just want to replace the number of 99999.00 with NaN but not with an Empty Field) , But when i want to use it manually from Table converted from Readmatrix by using this code :

[namafile,direktori]=uigetfile({'*.txt', 'Text-files (*.txt)';'*.sec', 'Seconds-Files (*.sec)';'*.min', 'Minutes-Files (*.min)'},'Load Data Magnet LEMI Non IAGA Non RAW (Menitan / Detikan)', 'Multiselect','on');
      cd(direktori); 
      full = fullfile(direktori,namafile);
      namafiles = cellstr(sort(namafile)); %Mensortir Kombinasi Data Yang Dipilih secara acak (PENTING)
      nfiles = size(full,2);
      f = cellstr(full);
      file = length(f);
      
opts = delimitedTextImportOptions('Delimiter',' ',...
      'VariableNames', {'VarName1','VarName2','VarName3','VarName4','VarName5','VarName6','VarName7','VarName8','VarName9','VarName10','VarName11','VarName12','VarName13'},...
      'WhiteSpace', '');
       opts.MissingRule = 'fill';
  
       Bs = readmatrix(full, opts);
       Bfix = cell2table(Bs);
       tabs = table(Bfix.Bs1, Bfix.Bs2, Bfix.Bs3, Bfix.Bs4, Bfix.Bs5, Bfix.Bs6, Bfix.Bs7,...
               Bfix.Bs8, Bfix.Bs9, Bfix.Bs10, Bfix.Bs11, Bfix.Bs12, Bfix.Bs13);
       tabss = varfun(@cell2mat,tabs);
       B = standardizeMissing(tabss,{'99999.00', 'N/A'});
       data = B(:,{'cell2mat_Var1','cell2mat_Var2','cell2mat_Var3','cell2mat_Var4','cell2mat_Var5','cell2mat_Var6','cell2mat_Var7','cell2mat_Var8',...
           'cell2mat_Var9','cell2mat_Var10','cell2mat_Var11','cell2mat_Var12','cell2mat_Var13'});
       tahun = str2num(data.cell2mat_Var1);
       bulan = str2num(data.cell2mat_Var2);
       tanggal = str2num(data.cell2mat_Var3); 
       jam = str2num(data.cell2mat_Var4);
       menit = str2num(data.cell2mat_Var5); 
       detik = str2num(data.cell2mat_Var6); 
       x = str2num(data.cell2mat_Var7); 
       y = str2num(data.cell2mat_Var8); 
       z = str2num(data.cell2mat_Var9); 
       DOF = str2num(data.cell2mat_Var13);
       Dlemi = datetime(tahun,bulan,tanggal,0,0,0,'TimeZone','Asia/Jakarta', 'Format', 'yyyy-MM-dd');
       D1 = datetime(tahun,bulan,tanggal,0,0,0,'TimeZone','Asia/Jakarta');
       DOHSlemi = sqrt((x.^2)+(y.^2));
       DOFSlemi = sqrt((x.^2)+(y.^2)+(z.^2));

The result of variable B (StandardizeMissing) is :

cell2mat_Var1    cell2mat_Var2    cell2mat_Var3    cell2mat_Var4    cell2mat_Var5    cell2mat_Var6    cell2mat_Var7    cell2mat_Var8    cell2mat_Var9    cell2mat_Var10    cell2mat_Var11    cell2mat_Var12    cell2mat_Var13
    _____________    _____________    _____________    _____________    _____________    _____________    _____________    _____________    _____________    ______________    ______________    ______________    ______________

        2019              07               14               00               00               59                              2027.10         -14498.37         35956.37          0.056406         -0.383280          38769.36   
        2019              07               14               00               01               59            35898.87          2027.14         -14498.32         35956.06          0.056408         -0.383282          38769.05   
        2019              07               14               00               02               59            35898.79          2027.35         -14498.38         35955.99          0.056414         -0.383284          38769.02  

Look at the cell2mat_Var7 row 1, its should be placed with NaN instead of " " or an empty field like that.....

The total amount of rows from each variable at the tables is should be 1440, but because using StandardizeMissing, the cell2mat_Var7 only contain with 1439 (missing 1 field). And because of that, the code become error at DOHSlemi :

        Matrix dimensions must agree.

Error in lemi2comp (line 56)
       DOHSlemi = sqrt((x.^2)+(y.^2));

So, how to use standardizeMissing function correctly to create NaN data instead of " " or an empty field by using a Table input?

This is the syntax of standardizeMissing at https://in.mathworks.com/help/matlab/ref/standardizemissing.html#bty9pdr-1-indicator :

B = standardizeMissing(A,indicator)
B = standardizeMissing(A,indicator,'DataVariables',datavars)

Thank you Community, i hope you can understand what the purpose of my question....

Thank you so much and please help me in finding the solution (i had already attached my data).... /.\ /.\ /.\

Answer 1

[Steven Lord]

My guess is that the data that was read into that 7th variable in your table was not numeric data but included data that MATLAB interpreted as text data. In that case the appropriate missing value for that variable is empty text, not NaN.

Can you show the output of this command?

class(B.cell2mat_Var7)

[Tyann Hardyn]

Alright, Sir @Steven Lord

the ouput is :

class(B.cell2mat_Var7)

ans =

    'char'

Iam sorry for late answer.... Its a char data type, Sir...

[Steven Lord]

Do you want that data to be stored as text (in which case blank text is the correct missing value) or as numbers (in which case NaN is the standard missing value)? If the latter you will need to convert the text into numbers using something like str2double. Alternately modify your importing code to specify that you want that variable in the table to be imported as double not char.

[Tyann Hardyn]

@Steven Lord.

Yes, Sir, i want the data to be in a form of NaN as the output, so the data input should be a Number? Alright, Sir, so i must to create a condition that the imported data is in a number (double).