just assume it's an integer then.

start with an example in pure math. show every step.
then write down what logic /conditions you used between each step.
then use variables instead of numerical values.
then convert to pseudocode.
then convert pseudocode to matlab

Did you do these steps yet?

Why the assumption? There's nothing special about the fractions, just keep iterating over the remainder until you either get an exact result or get to some maximum string length that you decide upon.

rough pseudo code:

result = "1232" //whatever whole part in string format
remainder = 0.123 //whatever the decimal fraction was
maxSteps = 20 //max fraction length in base 12 to approximate the fraction in base 10

for (int step = 0; step < maxSteps AND remainder != 0; step++) {
nextDigit = remainder * 12
result += nextDigit.toInteger().toString()
remainder = nextDigit - nextDigit.toInteger()
}


This post was edited by russian on Sep 18 2015 03:15pm