Even of Odd?

Im Looking for a math function that will give me a yes or no answer ( 0 or 1 ) as to whether or not any given number is even.

Mod

Is Mod what you are looking for ?

[email protected] said:

im looking for a math function that will give me a yes or no answer ( 0 or 1 ) as to whether or not any given number is even.

Click to expand...

I don't know what type of number you are looking at, or what type of system you are using. But IsOdd() should probably work

Mod is a very cpu intensive function. This is faster.
(Number AND 1) <> 0
This returns a true or false.

If the PLC being used allows to access numbers by bits, just check the least significant bit.

If your plc allows you to refer to integers at bit level then you can simply use the status of the least significant bit of your number:

0 = even
1 = odd

Look at the binary.

1 = 0001 - odd
2 = 0010 - even
3 = 0011 - odd
4 = 0100 - even
5 = 0101 - odd
6 = 0110 - even
7 = 0111 - odd

Do you see a pattern?

Check the least significant bit. If its a 1 then the number is odd.

In ladder

 MyInteger.0                IsOdd

----] [---------------------( )--

PS. you can also use a similar trick to see if a number is negative by checking the most significant bit; if it is a 1 then the number is negative.

An interesting original question - and my first thought was "he has to be asking about integers" (i.e. whole numbers).

Then the thought occurred, "If 4.000000 is even, is 4.000001 even or odd, or neither" ?
Can the concept of even-ness be applied to floating point numbers, taking into account the degree of imprecision of them?


EDIT : As always, Wikipedia provides the answer... even-ness applies only to Integer Numbers

http://en.wikipedia.org/wiki/Even_number

I like threads like this one. I tried this out on our system. If number is a INT or DINT I must convert the DINT to a DWORD first. According the the IEC specification I can't check bits of an integer. I tried
DINT_TO_DWORD(iNumber).0
but that didn't work but I think it should. I will bug the compiler writer about this.
I had to
wNumber:=DINT_TO_DWORD(iNumber);
Then check bit 0.
wNumber.0

Peter Nachtwey said:

I like threads like this one. I tried this out on our system.

Click to expand...

Which system is that Peter?

I find it hard to believe that any system would make it hard or impossible to access bit status of an integer when the system is dealing with that integer as a series of bits anyway!

It is only harder than it needs to be.

burnerman said:

Which system is that Peter?

I find it hard to believe that any system would make it hard or impossible to access bit status of an integer when the system is dealing with that integer as a series of bits anyway!

Click to expand...

I should have said our product. We are following the IEC specification where it makes sense. The IEC speficiation allows DWORD.0 but not DINT.0.

Customers want the IEC spefication but I really wonder if they know about all the stupid little things the IEC specificaitons makes difficult and it is so wordy.

This works
Odd:=(iTemp and 1)<>0; // Even this is a violation......., I mean extension of the specification. It should be
Odd:=(DINT_TO_DWORD(iTemp) and 1)<>0; // That is a lot of typing to do so little. What a pain.
This does too
Odd:=wTemp.0;
but not this
Odd:=iTemp.0;

I also hate the DINT_TO_DWORD() function. Too long and does nothing. I prefer simply DWORD(). This would convert the agrument to a DWORD whether it is a DINT or REAL.
Then I could type
Odd:=DWORD(iTemp).0;
This is less wordy but still keeps things type safe.

burnerman said:

Which system is that Peter?

I find it hard to believe that any system would make it hard or impossible to access bit status of an integer when the system is dealing with that integer as a series of bits anyway!

Click to expand...

That is actually a bit (NPI) backwards. The processors in PLC's, computers, well, everything actually deal with single bits as whole words (8/12/16/32/64/any other bit length is dependent on the system).
I can't think of any processor on the market (though I don't know all of them) that have bit level addressing.

For that matter, many processors don't have any direct access to the bits of a word, requiring the use of bitwise logical instructions (and, or, xor).