Finding if a particular value is found in a range of arrays?

Hi My range of arrays is :

array_fl =

       19950       20000       20050
       20950       21000       21050
       21950       22000       22050
       22950       23000       23050
       23950       24000       24050
       24950       25000       25050
       25950       26000       26050
       26950       27000       27050
       27950       28000       28050
       28950       29000       29050
       29950       30000       30050
       30950       31000       31050
       31950       32000       32050
       32950       33000       33050
       33950       34000       34050
       34950       35000       35050
       35950       36000       36050
       36950       37000       37050
       37950       38000       38050
       38950       39000       39050

I want to know if a number is found in this particular range, if yes, pick up its middle value else take the number as it is.

find below the code i wrote

num_fl=20;
array_fl=[];
initial_range=19950:50:20050;
% generate a range of flight level
for i=1:num_fl
  array_fl(i,:)=initial_range;
  initial_range=initial_range+1000;
end
%%getting the correct altitude
FL=[];

%example
Alt_greater_than_20000=[20960 27030 22800 29000]
length_array=length(Alt_greater_than_20000);
% for i=1:length_array
for j=1:num_fl
  for i=1:length_array
    if Alt_greater_than_20000(i) >= array_fl(j,1) & Alt_greater_than_20000(i) <= array_fl(j,3)
      FL(i)= array_fl(j,2);
    else
      FL(i)= Alt_greater_than_20000(i);
    end
  end
end    
end

find also my code. Can i know where i went wrong?

[Stephen23]

Rather than giving us broken code, it would be much easier if you actually clearly explained exactly what you are trying to achieve, and give example data of both input and output values that we can try out, and explain why those are the output values.

We know MATLAB quite well, so we can figure out some code to help you if you explain what it is that you are trying to achieve.

[yashvin]

For example, given that i have an array of numbers

Alt_greater_than_20000=[20960 27030 22800 29000]

For each number in the array, i want to know in which particular row of array_fl it lies. If this, i want to pick its middle value of the same row and store it. If it does not lie, i simply want to store it and continue with the next number.

For example for number 20960, i expect 21000, as it is located in the 2nd row of array_fl which is (20950 21000 21050).

for 27030, save it as 27000 (as it lies in range 26950 27000 27050)

for 22800, save it as 22800 (as it does not lie in any range)

for 29000, save it as 29000 (it lies in the range)

I expect the corresponding numbers to be [21000 27000 22800 29000]

I have also attached the M file of how i tried to do it.

[Purushottama Rao]

The problem in your code is associated with the no of comparisons made. Say for example first element 20960 is compared with 20 rows of the array_fl. Conditon is met in the second row and you are changing the Fl element corresponding to it. But during the next iteration, 20960 is compared against 3rd row and the condition is not met, therefore you are reverting again FL element to 20960. That is why you had a wrong result.

Answer 1

[Walter Roberson]

length_array=length(Alt_greater_than_20000);

for K = 1 : length_array
  fl_in_range = array_fl(:,1) <= Alt_greater_than_20000(K) & Array_greater_than_20000(K) <= array_fl(:,3);
  if any(fl_in_range)
    FL(K) = array_fl(fl_in_range,2);
  else
    FL(K) = Array_greater_than_20000(K);
  end
end

[yashvin]

What is Array_greater_than_20000? I am getting undefined...

[yashvin]

Changing it to Alt_greater_than_20000..let me try

[yashvin]

Yes it works!!! Thanks! The main thing was about "any". Thanks a lot!!!

[Guillaume]

There is actually a latent bug in Walter's code because of the use of any. If for some reason there happens to be more than one row of array_fn that matches, then any will return true and the following line, the assignment to Fl(K) will fail as it will try to assign an array to a scalar.

A safer approach would be to use sum. If it's 0, no row matches. If it's 1, exactly one row matches, if it's more than one, more than one row match:

switch sum(fl_in_range)
   case 0   %no match
       Fl(k) = Alt_greater_than_20000(K);
   case 1   %exactly one match
       Fl(k) = array_fn(fl_in_range, 2);
   case 2   %more than one match
       warning('more than one match found, using first one')
       Fl(k) = array_fn(find(fl_in_range, 1), 2);
end

Alternatively, add code to check that the ranges in fl_in_range do not overlap.

[yashvin]

I have one more question. If i want to eliminate the table ie the array_fl as the pattern will always be

19950 20000 20050

20950 21000 21050

...

How can i do it? It is just increment of 1000 between the rows. And if the value satisfies the range, it is only (19950+20050)/2 that i need to assign

[yashvin]

Anyone there? Or should i create a new post for this clarification?

[Walter Roberson]

m = mod(Alt_greater_than_20000, 1000);
need_round = m <= 50 | m >= 950;
Alt_greater_than_20000(need_round) = 1000 * round(Alt_greater_than_20000(need_round)/1000);

[yashvin]

Thanks!