String Comparison giving weird results, false for a true case.

Hi, I am writing code to take data from a large matrix(Measurements) and to break it into arrays of its subcomponents, for example Temperature. For the most part its fully functional besides in the cases where the component of the character array measurements, has empty or missing values. To try find out why it was not working i added lines 3 and 4 so that i could compare them.

for j = 2:1:289
               Measurements(j,i)
               string(missing) 
               strcmp(Measurements(j,i),string(missing)) 
              if strcmp(Measurements(j,i),string(missing)) == 0
                   Temperature(j-1) = str2double(Measurements(j,i));
                elseif  strcmp(Measuremnts(j,i),string(missing)) ==1
                    break
                end
end
_________________________________________________________________________________________________________________________________
ans = 

    "20.494556"

ans = 

    <missing>

ans =

  logical

   0

ans = 

    <missing>


ans = 

    <missing>

ans =

  logical

   0
__________________________________________________________________________________________________________________________________________

so those are a set of outputs, for some reason the code is giving a positive response even when the strings are the same, in which case it should break out of the loop but instead it records the missing values in the arrays as well.

How can I fix this? Why is it giving a are both cases in the string comparison given as , when they are clearly different?

Thanks.

[Geoff Hayes]

Joachim - what can you say about the data type for the measurement? It is clear that

string(missing)

is a string array (but I thought that you would have needed to add single quotes around 'missing' ??) but how about Measurements(j,i)? What does

class(Measurements(j,i))

return? Is it a string array or a character array? Maybe it doesn't matter (my version is so old that I don't have the string function...)

Answer 1

[Steven Lord]

If you're looking to find and/or fill missing data take a look at the ismissing and fillmissing functions.

rng default
A = randn(10, 1);
A(abs(A) > 0.75) = missing;
B = fillmissing(A, 'previous');
C = fillmissing(A, 'next');
D = fillmissing(A, 'linear');
miss = ismissing(A)
results = table(A, miss, B, C, D) % for ease of comparing the fill methods