import csv using textscan but skip lines where column value meets user criteria

I am wanting to import a csv using textscan where rows will be skipped if a column value meets a certain criteria.

Interval,Layer,Value,Comment

1,1,-78,ok

2,2,-89,ok

3, 0,-56,want this line gone

4,4,-34,ok

I wish to

This would reat the whole lot into data. I wish to exclude the third record where Layer = 0. Can this be done as a one-liner without having to resort to line-by-line reading ?

data = textscan(fid,'%f%f%fs','Delimiter',',')

Answer 1

[KSSV]

As the data int he file are numbers..you can import/ load the data and then you can remove the unwanted numbers using inequalitites.

REad about csvread, load, importdata and ineqalitites.

[Tim]

no there are comment strings at the end, so not all numeric

Have found a solution by creating an index variable and applying to each variable in turn using a for loop. Reasonably efficient I guess but was thinking that there is a way to do this in a single line as part of the textscan function. .

idx = find(data{2}~=0); % locate all rows where second columne value is not zero

for i=1:length(data);

data{i}=data{i}(idx); % step through each column and apply idx to remove non-zero rows

end

[KSSV]

YOu can read a file line by line as below:

fid = fopen(myfile);
tline = fgetl(fid);
while ischar(tline)
    disp(tline)
    tline = fgetl(fid);
end
fclose(fid);

Answer 2

[Tim]

That's right you can read line by line but this is inefficient especially for large files. Was wanting to just have a criteria for the textscan function to exclude lines where a condition is met and do it all in the one line. Thanks anyway KSSV.

[KSSV]

YOu can textscan the entire file into cells using textscan. Fin the required string using strcmp or contains. Remove those lines and write the data back to file.