Adding more than one greater/less than condition to an 'if statement'

Simply trying to construct an if statement asking whether a longitude/latitude coordinate for an image is located within a rectangular section.

 for i = 1:number_of_images
    if (im_lat(i) > lat_D && im_lat(i) < lat_A) && (im_lon(i) > lon_A && im_lon(i) < lon_B)
       ...
    end
 end

Where A is the top left corner of the rectangle, B is the top right, and D is the bottom left. I'm curious if there is an easy way of revising this statement. There may also be functions that do this, such as inpolygon; however, I'm not sure how to integrate anything like this with my current format. I greatly appreciate any information you are able to provide.

Answer 1

[dpb]

inpolygon can do it but you've then got to return both output variables and do the set difference between in and on to get those inside as it returns inclusive of boundary and your test is written as exclusive. In addition, it's more verbose to define the vertices of the square than write the explicit test value.

I use a little helper function that is simply "syntactic sugar" to simplify such that lets you write the above as

if iswithinex(im_lat(i),lat_D,lat_A) & iswithinex(im_lon(i),lon_A,lon_B)

where iswithinex is

>> type iswithinex
function flg=iswithinex(x,lo,hi)
% returns T for values within range of input
% SYNTAX:
%  [log] = iswithinex(x,lo,hi)
%      returns T for x between lo and hi values, exclusive

flg= (x>lo) & (x<hi);
>>

Originally I began with iswithin(x,lo,hi,flg) where the flag was for [in|ex]clusive but couldn't ever remember which way had defined it so I ended up with two specific versions. I suppose probably the way would have been to have made the flag input the string instead of a logical; in fact, I may try that variation and see how I like it... :)

[keegswegner]

Thank you for answering. I believe it works as intended; however, I left the output of the function as the default logical.

[dpb]

Say what??? I don't understand the comment at all...

[keegswegner]

I simply mean I set the outcome of the function to be either 1 or 0, where 0 is true and 1 is false. Therefore my updated if statement:

 if iswithinex(im_lat(i),lat_D,lat_A) == 0 & iswithinex(im_lon(i),lon_A,lon_B) == 0

[dpb]

Absolutely wrong idea! The function as written returns a logical which is, by definition 0/1 and the test against explicit numeric value is pointless as all that does is replace one T/F value with another.

If you mean you modified the function to actually return 0 for true, that violates every known programming convention in existence...a verry bad idea,indeed.

Demo:

    >> iswithinex(0,-1,1)
  ans =
    logical
     1
  >> iswithinex(2,-1,1)
  ans =
    logical
     0
  >> if iswithinex(0,-1,1),disp('true'),else,disp('false'),end
  true
  >> if iswithinex(2,-1,1),disp('true'),else,disp('false'),end
  false
  >>

From the Matlab help files...

>> help true
true   True array.
  true is short-hand for logical(1).
  true(N) is an N-by-N matrix of logical ones.
  true(M,N) or true([M,N]) is an M-by-N matrix of logical ones.
  ...
  true(N) is much faster and more memory efficient than LOGICAL(ONES(N)).
  ...

IF works on the expression--

>> help if
 if Conditionally execute statements.
    The general form of the if statement is

         if expression
           statements
         ...   

      The statements are executed if the real part of the expression 
      has all non-zero elements.
      ...

Thus in writing

if something==0

something==0 is _expression_ and that, if true returns (drum roll, please!) a logical 1.  Might as well just used the logical value instead and saved an extra test.  And, *NB* that the block is evaluated when the expression evaluates to 1, not 0.

Keep with convention; various languages have different choices in whether TRUE (~FALSE) is one or minus-one but I don't believe there's a single one that doesn't treat FALSE-->0

[keegswegner]

Ok, thank you for clarifying, I will look into this.

[dpb]

You're welcome...not following convention on zero for false will lead to much confusion for anyone else trying to follow your code and likely also eventually lead to logic errors as you'll continuously be fighting built in Matlab convention for every logical function there is--and there are a lot of 'em! :) Look at all the ISxxx() family for example.