Use a simpler expression for particle energy for reduced diagnostics

[dpgrote]

For the reduced diagnostic calculating the particle energy, this uses a simpler expression that does not have any approximations and is accurate for all values of beta.

The Taylor series expansion that was being used was very good, which a peak error of 1.e-12 at beta = 0.09. But the new expression has no error (beyond round off) and is much simpler.

Is there particle energy being calculated anywhere else in WarpX that could use the same simpler expression?

[lucafedeli88]

A very good solution! Thanks a lot for this PR!

I've quickly skimmed through the code and I haven't found other places were this expression could be used (but I haven't performed an exhaustive search).

[ax3l]

Naice!