write table into txt file using writetable returns error

i have this in my main, and its return a table called 'step2_result_euclidean' contains a table. This function may be used fews time and each being called i want to save it into a txt file for monitoring

   for ccc=1:k
        step2_euclidean = kira_euclidean(data,step1_result_centroid,ccc);
        result_step2=array2table(step2_euclidean);
        step2_result_euclidean(:,ccc)=result_step2(:,end) ;  
   end
   push_result(step2_result_euclidean);

This is my push_result function code:

function[]= push_result(data_to_be_push)

writetable(data_to_be_push,'result.dat''Delimiter',' ')

However, i get this error:

Error using writetable (line 124)
Wrong number of arguments.

How to fix this error? TQIA

Answer 1

[Cris LaPierre]

It looks like you have a syntax error when you call writetable. You need to separate each input with a comma. You are missing a comma between your file name and the name of your Name-Value pair. Try this.

function[]= push_result(data_to_be_push)

writetable(data_to_be_push,'result.dat','Delimiter',' ')

[Khairul nur]

tq for correcting my syntax error, the code is run..

however, the txt file store only the latest table generated by the loop..it suppose to store table generated by each loop

control_loop==0 
while control_loop<10 

   for ccc=1:k
        step2_euclidean = kira_euclidean(data,step1_result_centroid,ccc)
        result_step2=array2table(step2_euclidean)
        step2_result_euclidean(:,ccc)=result_step2(:,end)    
   end
   push_result(step2_result_euclidean);
   control_loop=control_loop+1
end
   

[Walter Roberson]

Is each iteration to store to a different file, or should each iteration to append to the end of the file?

[Khairul nur]

each iteration to append in the same file

[Walter Roberson]

writetable() has an writemode append option for text files.

[Khairul nur]

sadly not supported by R2019b as it is my version :( Correct me if i am wrong

i edit my code become :

function[]= push_result(data_to_be_push)

data=data_to_be_push
writetable(data_to_be_push,'result.dat','WriteMode','append ')

has error:

Error using writetable (line 124)
Invalid parameter name: WriteMode.

Error in push_result (line 4)
writetable(data_to_be_push,'result.dat','WriteMode','append ')

Any other solution for my problem?

[Cris LaPierre]

If you can't update your version, then perhaps you should look into low-level file i/o. This is actually what writetable is using to write the data. Use fopen with the 'a' permission to append data to the file. You then use fwrite to write the data, one line at a time, to the file. Then use fclose to close the file.

Here is a template (untested). It assumes the table only contains numeric data.

function[]= push_result(data_to_be_push)
fid = fopen('result.dat','a');
fwrite(fid,data_to_be_push{:,:},'double');
fclose(fid);
end

[Walter Roberson]

fwrite() is for writing in binary, which the user probably does not want to do (they used 'delimiter' option, which is only for text.)

For text writing, you would want fprintf. Assuming that data_to_be_push is a table of numeric data, then

function[]= push_result(data_to_be_push)
if isempty(data_to_be_push); return; end

filename = 'result.dat';   %but in which directory??
[fid, message] = fopen(filename,'a');
if fid < 0
    error('failed to open file "%s" because: "%s"', filename, message);
end

fmt = [repmat('%.18g ', 1, size(data_to_be_pushed,2)-1), '%.18g\n'];

fprintf(fid, fmt, data_to_be_push{:,:}.'); %transpose is important!

fclose(fid);
end

Here I use a .18g format, whereas you are probably more accustomed to seeing .16g . Roughly 1/5 of floating point numbers need .17g instead of .16g to fully reproduce, and I recently saw argument that .18g is needed in some borderline cases (that end in multiple 9's). I am sure that .17g is needed, not certain that .18g is needed.

[Cris LaPierre]

Ha. Well, I did say it was untested. Yes, fprintf is a better choice.

[Khairul nur]

tq Cris and Walter, i appreciated ur reply :) However i still facing error. Please help me to solve this.

Btw walter, i trying to understanding on 18g and 17g and all as i the code as below error:

fmt = [repmat('%.16g ', 1, size(data_to_be_pushed,2)-1), '%.16g\n'];

since i try to play around and changing your code using '16g' as below code:

function[]= push_result(data_to_be_push)

if isempty(data_to_be_push); return; end
filename = 'C:\Users\Khairul Nurmazianna\OneDrive\Desktop\FCM\result.dat';   %but in which directory??
[fid, message] = fopen(filename,'a');
if fid < 0
    error('failed to open file "%s" because: "%s"', filename, message);
end
fmt = [repmat('%.16g ', 1, size(data_to_be_pushed,2)-1), '%.16g\n'];
fprintf(fid, fmt, data_to_be_push{:,:}.'); %transpose is important!
fclose(fid);
end

This is the example of data_to_be_pushed (table) looks like, each iteration will produce the same size of table.

1.46995037518577	0.685711409618569	0.627605582262098
1.80125654902813	2.11239700592433	2.40486047110991
4.96690963594932	5.82573980407720	6.29348791553382
10.1325627228705	11.5390826022301	12.1821153599577
0.884083765656048	0.642946391309453	0.498325558636135
4.38104302641960	5.78297478576809	6.16420789190786
2.29821715612633	2.60018137300034	2.36904553501017
1.46387024304751	2.31352417115322	2.25767297943408
9.54669611334078	11.4963175839210	12.0528353363318
3.22147007163667	2.44181922608056	2.64551305031197

[Walter Roberson]

fmt = [repmat('%.18g ', 1, size(data_to_be_push,2)-1), '%.18g\n'];

I had an error in the variable name, as 'data_to_be_pushed' is more natural than 'data_to_be_push'

[Khairul nur]

hihi..ok..the variable 'pushed'..i overlook on this as trying so hard to understand repmat function 18g or 17g..and its work! ur are great in matlab Mr. Walter and helpfull too.. :)