Turn a row vector with -99 separating data sets into a matrix

Some experience with Matlab, but a complete newcomer to the programming aspect. Trying to teach myself a bit with the Stormy Attaway book.

One of the problems in the while loops section is to read data from a file containing a single row vector of data with data sets separated by a flag -99, create a new vector that only has data values up to but not including the -99, and plot the values. That code looks like this:

load experd.dat

i = 1;
while experd(i) ~= -99
    newvec(i) = experd(i);
    i = i + 1;
end

plot(newvec,'ko')
xlabel('Reading #')
ylabel('Weight(pounds)')
title('First Data Set')

I wanted to modified the code a bit to plot the values after the -99 as well so I wrote the following:

load experdata.txt

i = 1;

while experdata(i) ~= -99
    newvec1(i) = experdata(i);
    i = i+1;
end

plot(newvec1, 'o')
xlabel('Reading #')
ylabel('Weight (pounds)')
title('First data set')
hold on

i = i+1;
[m,n] = size(newvec1);
[r,c] = size(experdata);

while i <= c
    newvec2(i-(n+1)) = experdata(i-(n+1));
    i = i+1;
end

plot(newvec2, '*')
xlabel('Reading #')
ylabel('Weight (pounds)')
title('First data set')

But, then I thought it would be nice to modify the code to handle a case in which there are an unknown number of data sets separated by -99s. I started to write the following, which kind of works but doesn't give me exactly what I would like.

    load experdata.txt

    [m,n] = size(experdata);

    i = 1;
    x = 1;

    while i <= n
        if experdata(i) ~= -99
            newvec(x,i) = experdata(i);
            i = i+1;
        else
            i = i+1;
            x = x+1;
        end
    end

The problem is that you get a matrix that looks like

1 2 3 4 0 0 0 0

0 0 0 0 5 6 7 8

where there original data would have been 1 2 3 4 -99 5 6 7 8,

whereas I would prefer

1 2 3 4

5 6 7 8

The problem is that I can't figure out how to reset the index when the code passes over the -99.

Any insight would be greatly appreciated. Thank you. Nate

[Image Analyst]

You'd get more attention if people could try your code. Attach experdata.txt.

Answer 1

[Andrei Bobrov]

out = zeros(size(experdata));
l0 = experdata == -99;
ii = find(l0)+1;
out([1,ii]) = 1;
out = cumsum(out(~l0));
out_main = accumarray(out(:),experdata(~l0),[],@(x){x});