Hi - Firstly, I don’t think it’s possible and secondly, I think even if you do implement those tiny / large numbers, you’ll run into all sorts of floating point errors later on.
Much better to find a ‘clever’ way to simulate the same concept / experience…
How can I use float numbers in scientific format 12.345e+12 in blueprint? Unreal Engine Editor don’t store this numbers as literal values, for example gravity constant 6.67e-11. If I try to input this number then it reset to 0.0.
How do you implement formula: F = GMm/r^2 - where G ~ 1e-11, M ~ 1e+26, m ~ 1e+2 and r ~ 1e+8 and value changing? What error? Why? It simply formula, that implementation in simply gravity simulator which creates students in 1st year in university) And all works well. I do not see problems in this. In all applications implements setting float number in edit field, but Unreal couldn’t it.
UE4 don’t decide how float are stored, this is defined by CPU and it instruction set which usally is follows IEEE 754 standard, you can read about it here
Blueprints use single precision float types stored in 32-bits it most common float type used, in C++ you can also use double precision floats (64-bits) which let you store floats with more precise fractions, but those are not supported in blueprint, only in property editor
Remember you dealing here game engine, it was not made for scientific accuracy, it simply use default support of float in hardware. The precision dealing is done fully by CPU.
Unreal isn’t Matlab. There may be a way of doing this in UE, I’m just not aware of it…
Correct me if I’m wrong ( anyone ), but UE isn’t for crunching numbers, it’s for producing a visible end result.
How you do that doesn’t really matter…
I am needn’t very high precision, I need simply float number 123.456E+26 and enter it in edit field in blueprint. In C++ mode I use float number, but blueprint not supoort displaying float number in exp format 123.456E+26. Simply number, nothing more. I will try modify code and recompile UE4 from sources, because this situation very disillusion)).
It’s done)) I modified source code and create pull-request on github. Now, I am waiting approvement. Maybe in one of next releases will have support scientific numbers.