Dissimilar definition of EQUAD in Enterprise propagating here

[Hazboun6]

The definition of EQUAD, in the presence of EFAC, is not the same in Enterprise as it is in the 3 major timing software packages, Tempo, Tempo2 and Pint. In the timing packages EQUAD is inflated by EFAC, in the same way as the TOA errors. In Enterprise and in TempoNest they are not. The origin of the Enterprise convention is from TempoNest.

PR 315 makes these changes in Enterprise and the same needs to be done for the uses of WN here. The decision was made to match the timing packages in our default behavior.

[Hazboun6]

@AaronDJohnson and @paulthebaker do you have thoughts about whether we should allow the TempoNest style behavior and how we should flag that? I think we probably need to keep the old convention. How about something like wn_convention='tempo', and the other possible flag would be 'TN'?

[paulthebaker]

Your suggestion sounds good.

I prefer boolean flags when possible over strings. They're harder to screw up. Since the tempo convention will be the default, maybe do a 'wn_temponest=Falseorwhitenoise_TN=False` instead. I don't feel super strongly about it, though.

[siyuan-chen]

I agree with Paul that a boolean wn_tn=False is a safer option.

Related to conventions between the packages: The DM amplitude in temponest and enterprise also differ and can also be fixed by an option to use either, keeping the current one as the default. The temponest DM amplitudes are ~1.64 larger in log10 or ~43.4 times larger in linear scale.