Rounding decimals matlab manipulates

I need to write, in 1 code line, given a number (pos, neg, int, float), how do I round it without preset functions like fix, round, etc.?

[Walter Roberson]

How would you do it in multiple lines of code?

hint 1: watch out for negative

hint 2: log10 to determine the number of decimal places it already has

hint 3: watch out for 0

hint 4: watch out for integers that are exact powers of 10: for example, log10(100) = 2 exactly, log10(99.999) = slightly less than 2.

hint 5: x * (1-eps) for positive x is always less than x

[Stephen23]

yuval ohayon's "Answer" moved here:

Well thank for the hints Accualy in one code line zerp loops. Smone can help?

[Guillaume]

@Yuval,

a) Use Comment on this question rather than starting an answer

b) proofread what you read. The above is incomprehensible.

[Jan]

@yuval ohayon: "without the preset function like fix,round" is not clear enough. Is "rem" and "mod" allowed or not? If so, simply use "rem(..., 1)".

[Walter Roberson]

Is floor() permitted?

[yuval ohayon]

Im not familier with rem function,floor isn permitted. For my understanding it must be with mod func,but how i find if the numbet is upper .5 or down .5

[yuval ohayon]

Someones help please

[Steven Lord]

In general people on MATLAB Answers don't post answers to homework questions or questions that sound like homework. If you show what you've done to try to solve the problem on your own and ask a specific question about where you're having difficulty you may receive some more specific suggestions.

Of course, if this isn't a homework problem, just use the round function or the other rounding functions included in MATLAB.

[Image Analyst]

mod() is also a built in function like round(), fix(), int32(), ceil(), floor(), etc.

http://www.mathworks.com/matlabcentral/answers/38787-what-can-be-programmed-without-any-built-in-functions

Answer 1

[Jan]

Try it:

x = 13.3
y = mod(x, 1)
z = 13.6
y = mod(x, 1)

Now you should see how to use the output to decide, if the rest of the division by 1 is lower or higher than 0.5.

[yuval ohayon]

Well ,how i make condision with no loop x=input number Y=x-xmod(x,1)

Help

[Walter Roberson]

x - mod(x,1) + mod(x,1)>0.5

[Jan]

int64() rounds also. And you can use cast() to preserve the original class.