Support for Epsilon on double and float

[GooRiOn]

Currently we don't support Epsilon in extensions for double and float. The suggestion is that each method should have default parameter like:

(this IValitRule<TObject, double> rule, double value, double epsilon = 0.0d)

so user can specify it, but it's not required. Each method should then consider this number in the calculations. Need to applied for:

• IsGreaterThan()
• IsLessThan()
• IsEqualTo()
• IsPositive()
• IsNegative()
• IsNonZero()

[GKotfis]

Help still needed? I can pick up this one :)

[GooRiOn]

@GKotfis sure! Feel free to take this one :)

[GKotfis]

Before I change bunch of code I'd like to show my approach.
I used algorithm from https://floating-point-gui.de/errors/comparison/
Here are the changes for IsGreaterThan for Double
GKotfis@aad3d76
I haven't write new unit test so far - just checked that this change doesn't broke anything.
What do you think about that?

[GooRiOn]

I read the article and it looks good to me. Your implementation also looks good, however I think you should also consider handling Double.NaN values inside IsNearlyEqual method.