I don't know much about log and exp relationship. But, i am confused that, log is not used in softmax regression?
Features to Probabilities section
Exp is defined for all real values, while log isn't.
Any vector element equal or less than zero would lead to a numerical error or undefined behavior with the log function. exp does not have this problem.
To avoid the the undefined, you would either need to add a very small bias to all your elements. But how small should this epsilon be? What do you do with negative numbers. You may never expect a negative number but simply get it through a truncation error or coming very close to floating point representations (a value so small it ends up being represented as a small negative number).
Related
I have a program which calculates probability values
(p-values),
but it is entering a very large negative number into the
exp function
exp(-626294.830) which evaluates to zero instead of the very small
positive number that it should be.
How can I get this to evaluate as a very small floating point number?
I have tried
Math::BigFloat,
bignum, and
bigrat
but all have failed.
Wolfram Alpha says that exp(-626294.830) is 4.08589×10^-271997... zero is a pretty close approximation to that ;-) Although you've edited and removed the context from your question, do you really need to work with such tiny numbers, or perhaps there is some way you could optimize your algorithm or scale your numbers?
Anyway, you are correct that code like Math::BigFloat->new("-626294.830")->bexp seems to take quite some time, even with the support of use Math::BigFloat lib => 'GMP';.
The only alternative I can offer at the moment is Math::Prime::Util::GMP's expreal, although you need to specify a precision to it.
use Math::Prime::Util::GMP qw/expreal/;
use Math::BigFloat;
my $e = Math::BigFloat->new(expreal(-626294.830,272000));
print $e->bnstr,"\n";
__END__
4.086e-271997
But on my machine, even that still takes ~20s to run, which brings us back to the question of potential optimization in other places.
Floating point numbers do not have infinite precision. Assuming the number is represented as an IEEE 754 double, we have 52 bits for a fraction, 11 bits for the exponent, and one bit for the sign. Due to the way exponents are encoded, the smallest positive number that can be represented is 2^-1022.
If we look at your number e^-626294.830, we can do a change of base and see that it equals 2^(log_2 e · -626294.830) = 2^-903552.445, which is significantly smaller than 2^-1022. Approximating your number as zero is therefore correct.
Instead of calculating this value using arbitrary-precision numerics, you are likely better off solving the necessary equations by hand, then coding this in a way that does not require extreme precision. For example, it is unlikely that you need the exact value of e^-626294.830, but perhaps just the magnitude. Then, you can calculate the logarithm instead of using exp().
I want to display an Error message when the signal reaches a certain value or simply when it reaches 0 I've used the ifblock and Relational Operatorbut it does not work for me.
You're most likely checking if the signal is exactly zero, which with floating point arithmetic is almost always a very bad thing to do.
Rather, you want to check that the absolute value of the signal is less than some small tolerance. More than that, you almost certainly need to check if the average of the signal over the past n-time points (where you choose n) is less than the tolerance.
You might also consider using something like the Static Gap block from the Model Verification library.
I am trying to turn off denormal number support in matlab, so that basically any two computations that would result in a denormal number would instead just result in zero (DAZ, FTZ)
I've researched several sites include the one below, but I haven't found anything about doing this.
http://blogs.mathworks.com/cleve/2014/07/21/floating-point-denormals-insignificant-but-controversial-2/
I've never heard of such an option in Matlab. It would likely require deep manipulation of a lot of the floating-point math, effectively requiring a new datatype to be supported if this were to be an easily toggle-able option in Matlab. You could write your own mex C code to do this (more here and here) for an individual function.
And of course you can get something like this with one line of Matlab – here's an example:
a = [1e-300 1e-310 1e-310];
b = [1e-301 1e-311 1e-310];
x = a-b;
x(abs(x(:)) < realmin(class(x))) = 0;
where realmin is the smallest normalized floating-point number. However, the floating point math is still performed using the extended denormal/subnormal values in a. It's just the output that's clipped to zero.
Unless you're doing this for fun an experimentation, or possibly running code on an embedded platform, I'd really recommend against disabling denormals as a form of optimization. Instead, focus on why your values are so small and how you might rescale your problem to avoid the issue entirely.
I'm trying to write a basic digit counter (an integer is inputted and the number of digits of that integer is outputted) for positive integers. This is my general formula:
dig(x) := Math.floor(Math.log(x,10))
I tried implementing the equivalent of dig(x) in Ruby, and found that when I was computing dig(1000) I was getting 2 instead of 3 because Math.log was returning 2.9999999999999996 which would then be truncated down to 2. What is the proper way to handle this problem? (I'm assuming this problem can occur regardless of the language used to implement this approach, but if that's not the case then please explain that in your answer).
To get an exact count of the number of digits in an integer, you can do the usual thing: (in C/C++, assuming n is non-negative)
int digits = 0;
while (n > 0) {
n = n / 10; // integer division, just drops the ones digit and shifts right
digits = digits + 1;
}
I'm not certain but I suspect running a built-in logarithm function won't be faster than this, and this will give you an exact answer.
I thought about it for a minute and couldn't come up with a way to make the logarithm-based approach work with any guarantees, and almost convinced myself that it is probably a doomed pursuit in the first place because of floating point rounding errors, etc.
From The Art of Computer Programming volume 2, we will eliminate one bit of error before the floor function is applied by adding that one bit back in.
Let x be the result of log and then do x += x / 0x10000000 for a single precision floating point number (C's float). Then pass the value into floor.
This is guaranteed to be the fastest (assuming you have the answer in numerical form) because it uses only a few floating point instructions.
Floating point is always subject to roundoff error; that's one of the hazards you need to be aware of, and actively manage, when working with it. The proper way to handle it, if you must use floats is to figure out what the expected amount of accumulated error is and allow for that in comparisons and printouts -- round off appropriately, compare for whether the difference is within that range rather than comparing for equality, etcetera.
There is no exact binary-floating-point representation of simple things like 1/10th, for example.
(As others have noted, you could rewrite the problem to avoid using the floating-point-based solution entirely, but since you asked specifically about working log() I wanted to address that question; apologies if I'm off target. Some of the other answers provide specific suggestions for how you might round off the result. That would "solve" this particular case, but as your floating operations get more complicated you'll have to continue to allow for roundoff accumulating at each step and either deal with the error at each step or deal with the cumulative error -- the latter being the more complicated but more accurate solution.)
If this is a serious problem for an application, folks sometimes use scaled fixed point instead (running financial computations in terms of pennies rather than dollars, for example). Or they use one of the "big number" packages which computes in decimal rather than in binary; those have their own round-off problems, but they round off more the way humans expect them to.
I compute this simple sum on Matlab:
2*0.04-0.5*0.4^2 = -1.387778780781446e-017
but the result is not zero. What can I do?
Aabaz and Jim Clay have good explanations of what's going on.
It's often the case that, rather than exactly calculating the value of 2*0.04 - 0.5*0.4^2, what you really want is to check whether 2*0.04 and 0.5*0.4^2 differ by an amount that is small enough to be within the relevant numerical precision. If that's the case, than rather than checking whether 2*0.04 - 0.5*0.4^2 == 0, you can check whether abs(2*0.04 - 0.5*0.4^2) < thresh. Here thresh can either be some arbitrary smallish number, or an expression involving eps, which gives the precision of the numerical type you're working with.
EDIT:
Thanks to Jim and Tal for suggested improvement. Altered to compare the absolute value of the difference to a threshold, rather than the difference.
Matlab uses double-precision floating-point numbers to store real numbers. These are numbers of the form m*2^e where m is an integer between 2^52 and 2^53 (the mantissa) and e is the exponent. Let's call a number a floating-point number if it is of this form.
All numbers used in calculations must be floating-point numbers. Often, this can be done exactly, as with 2 and 0.5 in your expression. But for other numbers, most notably most numbers with digits after the decimal point, this is not possible, and an approximation has to be used. What happens in this case is that the number is rounded to the nearest floating-point number.
So, whenever you write something like 0.04 in Matlab, you're really saying "Get me the floating-point number that is closest to 0.04. In your expression, there are 2 numbers that need to be approximated: 0.04 and 0.4.
In addition, the exact result of operations like addition and multiplication on floating-point numbers may not be a floating-point number. Although it is always of the form m*2^e the mantissa may be too large. So you get an additional error from rounding the results of operations.
At the end of the day, a simple expression like yours will be off by about 2^-52 times the size of the operands, or about 10^-17.
In summary: the reason your expression does not evaluate to zero is two-fold:
Some of the numbers you start out with are different (approximations) to the exact numbers you provided.
The intermediate results may also be approximations of the exact results.
What you are seeing is quantization error. Matlab uses doubles to represent numbers, and while they are capable of a lot of precision, they still cannot represent all real numbers because there are an infinite number of real numbers. I'm not sure about Aabaz's trick, but in general I would say there isn't anything you can do, other than perhaps massaging your inputs to be double-friendly numbers.
I do not know if it is applicable to your problem but often the simplest solution is to scale your data.
For example:
a=0.04;
b=0.2;
a-0.2*b
ans=-6.9389e-018
c=a/min(abs([a b]));
d=b/min(abs([a b]));
c-0.2*d
ans=0
EDIT: of course I did not mean to give a universal solution to these kind of problems but it is still a good practice that can make you avoid a few problems in numerical computation (curve fitting, etc ...). See Jim Clay's answer for the reason why you are experiencing these problems.
I'm pretty sure this is a case of ye olde floating point accuracy issues.
Do you need 1e-17 accuracy? Is this merely a case of wanting 'pretty' output?
In that case, you can just use a formatted sprintf to display the number of significant digits you want.
Realize that this is not a matlab problem, but a fundamental limitation of how numbers are represented in binary.
For fun, work out what .1 is in binary...
Some references:
http://en.wikipedia.org/wiki/Floating_point#Accuracy_problems
http://www.mathworks.com/support/tech-notes/1100/1108.html