Power spec

[boryanah]

Change including TSC offset and minor bugs for LCV

[boryanah]

@lgarrison updating the quick thing for CV, but ran into classy compile issues (and I think the other tests got canceled alongside that).

[lgarrison]

There's a workaround for classy on main, can you merge main into this branch?

[boryanah]

Merged it, but there seems to be some issue now with Corrfunc -- not sure why it only shows up in this branch

[lgarrison]

Looks like it was just a transient issue, re-running the jobs fixed it.

[lgarrison]

@boryanah In using calc_power(), I realized we didn't have default values for most of the arguments, so I set some. Can you check that they look good? The only one that's not optional is Lbox.

This did have the side effect of changing the order of the arguments, so just be aware in case you have any code that's using calc_power() already (I changed all the places it was used in the repo).

[boryanah]

Thanks! The default values make sense for calc_power(). In describing the changes here, I forgot to say that ic_fields.py was optimized, too (everything now uses numba). There was also a mode counting bug, which doesn't matter in squaring k_x,y,z.

[boryanah]

I wonder if we should switch to passing k_bins and mu_bins directly in calc_power, but perhaps we can leave this for later.

[lgarrison]

I was thinking about changing the kbin/mubin arguments too, so I just went ahead and did it. Now they both accept either a number of bins or an array of bins.

[boryanah]

Ok, this looks great to me, so I'll just merge -- thanks.