Sorry your browser is not supported!

Have you ever written a statement like this: if n = 1 then m = 1 else m = 0 and then (assuming both are boolean) felt a bit silly for not writing m = n?
What about if n = 1 then m = 0 else m = 1? Would you feel equally silly (again assuming both are boolean) for not writing it as m = 0-n?
Did you know that you can use the results of relational comparisons to feed into further logic?: if (m>0) and (n=4) then print "Wha!" and that multiplication can be used like the logical AND, and sometimes even reduce the number of operations? (a*b is the same as a and b) Did you know that addition does the same for logical OR?

This is just a tiny peak into the world of mathematical conditions. Sometimes they are faster than conditional statements, sometimes they are slower, but overall I find it a lot of fun to try and condense something that would usually take several lines into one expression. It's become sort of a hobby within a hobby for me. Sometimes you can come up with code that's very useful by doing this, especially if the result of the condition is simply to assign a value.

I'll follow with some examples I think are particularly pleasing:

Perhaps the simplest, but nonetheless elegant, example I can think of is toggling a value between 1 and 0: n = 1-n.

A common problem, probably the first problem you face as a new coder, is to manage key presses; the challenge is usually to extract one instance of a key press from a key being held down. The common solution is called a "trip switch": because it only fires once until it is reset (in this case when the key is released) like a circuit breaker in a fuse box. The usual code is like so:
I used this for a very long time until discovering a mathematical solution: k = keystate(x) * (k + 1). The "+1" is very important here, it's what allows the switch to activate (k=1) in the first place (since on compile k=0), but upon the next test "+1" comes into play again and increments k. As long as key x is being held k will continue to count upwards, and so it can be used as a simple "trip" if k=1 and the incremental values can also be utilised, but when key x is released it pulls the rug under k and wipes its value.

Wrapping a values to stay within a certain range is another fairly common need. For angles we have WRAPVALUE but for different ranges it can be tricky for newcomers. Here's a mathematical solution for ranges with a lower limit of 0:
Let n equal the subject variable, Let b equal the upper limit
n = ((n < 0) + (n / b - int(n / b))) * b
and for any lower limit I added a few extra assignments to help the processor:
Let n equal the subject variable, Let a equal the lower limit, Let b equal the upper limit
n = n - a : b = b - a
n = ((n < 0) + (n / b - int(n / b))) * b + a

You can assign boolean results to a variable since they return basically as an integer 0 or 1. (edit: I see you did that below already)

m = (n > 0)



If a was 42 and b was 0, the first statement would not equal the second. I'm sure you see that, so maybe I misunderstood what you were implying.


Also, shorter condensed code is not always necessarily faster.


Your wrap method can be faster, but it can be slower in some cases. I ran a few tests and found for and number that's less than 3 times the bounding value, the repeat statement worked faster. But obviously, anything greater than that then your method is faster with a consistent time. I didn't try anything with the lower limit.

In Ansi C the "expr1 ? retTrue : retFalse" comes to mind.

string Answer = ( input == 'y' ? "Hell yeah!" : "Aww heck..." );

http://msdn.microsoft.com/en-us/library/ty67wk28.aspx

isnt this backards, multiplication gives logical OR

a=1
b=0
a*b=0

if either condition is not 0 it will still return true, both must not equal 0 to return false thus if a=0 or b=0

while addition gives logical AND

a=1
b=0
a+b=1

here both must be 0 to return true thus if a=0 and b=0

i tend to set toggle = 1 and then go on to use toggle = -toggle in boolean fashion. it's probably a wasteful habit but it is pretty simple/handy.

Not quite. Think about the truth table for AND and truth table for OR



If you multiple p and q you get the equivalent of AND and if you add p and q, you get the equivalent of OR.

Edit- the spacing didn't come out right.


Should that be m = abs(0-n)?

thats true if you are looking at the result value as the logical return, however the statement given was 'if a*b=0' not 'if a*b'

by testing the result value against 0 it flips

if 1*1=0 gives if 1=0 returns false so if 1=0 or 1=0 returns false
if 1*0=0 gives if 0=0 returns true so if 1=0 or 0=0 returns true
if 0*0=0 gives if 0=0 returns true so if 0=0 or 0=0 returns true

Oh I see what you're saying. I thought we were comparing to 0 and stopping there, not looking at it as a whole. Yep, you're right.

hi OBese87,

your equation is interesting, it could also be of interest to users DBClassic.
I often use the way:

for 3 or 4 keys, it is not long to write, but more keys, I can use an equation similar to yours:
k = keystate(x) | (2*(k&1))

there has only 4 results: 0,1,3,2
0 = long released
1 = short press
3 = long press
2 = short released

and the way is easy to put in a function:

You're right, I don't know why I added the "=0", changed.

@Ortu
You are right also!: "a*b = 0" equals "a=0 or b=0". Sorry for the confusion.

@Silverman
(You mean "2 = short release" and "3 = long press".) That's a nice one. It adds the ability to perform an action when the key is released, but sacrifices the counting functionality.