Analysis Packages

Analyses

Here is the full list of analysis functions that can be specified/performed in pyharm. These can be listed as arguments to pyharm-analysis.

Groups of particular related reductions performed over a GRMHD run.

Each function takes a dump object and a dict ‘out’ to fill when computing new variables. This allows collation in analysis.py for the pyharm-analysis script

Variables are organized by remaining independent variable, as described in ana_results.py

pyharm.ana.analyses.basic(dump, out, **kwargs)

Anything necessary for other functions: t, EH fluxes, whether to average

pyharm.ana.analyses.diagnostics(dump, out, **kwargs)

Energy ratios on the grid to gauge floor effectiveness, along with total floor hit and inversion flags

pyharm.ana.analyses.dynamo(dump, out, **kwargs)

Compare magnetization in the upper and lower hemisphere of EH, and at 5 r_g

pyharm.ana.analyses.efluxes(dump, out, **kwargs)

Total energy fluxes, recorded so that ~div-free steady state flux can be computed

pyharm.ana.analyses.gridtotals(dump, out)

Total energy and current, summed by shells to allow any cut on radius

pyharm.ana.analyses.jet_cut_lite(dump, out, **kwargs)

Compute jet powers with just the default cut from EHTC Paper V ‘19. These are the powers used in the table in that paper and the MAD Code Comparison ‘22

pyharm.ana.analyses.jet_cuts(dump, out, **kwargs)

Blandford-Znajek Luminosity L_BZ, calculated with a sampling of different cuts floated for EHTC V ‘19

pyharm.ana.analyses.jet_profile(dump, out, **kwargs)

Theta profiles of many (many) different variables at some particular radius. Used to gauge what is/isn’t jet and where it wanders

pyharm.ana.analyses.lumproxy(dump, out, **kwargs)

Luminosity proxy from GRMHD code comparison ‘19

pyharm.ana.analyses.madcc(dump, out, **kwargs)

Functions for MAD Code Comparison ‘22, nontrivial mandatory diagnostics. See that paper/doc for full descriptions

pyharm.ana.analyses.madcc_optional(dump, out, **kwargs)

Functions for MAD Code Comparison ‘22, optional diagnostics

pyharm.ana.analyses.omega_bz(dump, out, **kwargs)

Field rotation rate, for comparison against Blandford-Znajek prediction

pyharm.ana.analyses.omega_bz_advanced(dump, out, **kwargs)

A battery of different measurements of the Blandford-Znajek prediction of the B field rotation rate.

pyharm.ana.analyses.outfluxes(dump, out, **kwargs)

Outflowing portions of fluxes

pyharm.ana.analyses.pdfs(dump, out, **kwargs)

Probability density functions for certain variables. Even more likely than most analyses to be broken

pyharm.ana.analyses.print_divb(dump, out, **kwargs)

Just check divB. Used as a cursory check for restart file sanity

pyharm.ana.analyses.print_flags(dump, out, **kwargs)

Just check floor and failure flags. Used as a cursory check for run sanity

pyharm.ana.analyses.r_flux_profiles(dump, out, vars=('FM', 'FE', 'FL'), **kwargs)

Radial profiles of conserved mass, energy, angular momentum

pyharm.ana.analyses.r_profile_phi(dump, out, **kwargs)

Spherical and midplane magnetizations of radial shells, analogous to FM or FE for Phi_b

pyharm.ana.analyses.r_profiles(dump, out, vars=('rho', 'Pg', 'u^r', 'u^3', 'u_3', 'b', 'inv_beta', 'Ptot'), **kwargs)

Calculate Radial profiles, by averaging over phi and some portion of theta. Separate averages over the comparison “disk” portion and the rest of the domain, marked “notdisk”. Keeps time-dependent versions for averaging post-hoc

pyharm.ana.analyses.r_profiles_cc(dump, out, **kwargs)

Radial profiles of everything used in the MAD Code Comparison ‘22

pyharm.ana.analyses.rth_profiles(dump, out, vars=('betainv', 'rho', 'sigma', 'Theta'), **kwargs)

2D profiles in r/th, time-averaged. Default betainv, rho, sigma, Theta

pyharm.ana.analyses.th_profiles(dump, out, vars=('inv_beta', 'sigma'), **kwargs)

Calculate any full-theta profiles: 5-zone radial averages starting from ‘rTh’ (also averaged in phi). Keeps time-dependent versions for averaging post-hoc

Reductions

Should (when) you want to add your own analysis operations, there are quite a few integrals and other reduction operations defined in pyharm already. Note you can also find sample usage in pyharm.ana.analyses.

The general interface for reductions is (dump, var, options) | dump: FluidDump object, generally pre-slice. | var: either the name of a variable to compute over the necessary slice, or a pre-sliced & computed variable of the right shape.

Other options as described below. Certain zone arguments with defaults are still necessary when a reduction is otherwise ambiguous!

pyharm.ana.reductions.corr_length_phi(R, interpolate=True)

Correlation “length” (angle) given a correlation function of r,phi

pyharm.ana.reductions.corr_midplane(var, norm=True, at_i1=None)

Angular correlation function at the midplane, of an array representing a variable in spherical-like coordinates r,th,phi

pyharm.ana.reductions.corr_midplane_direct(var, norm=True)

Alternate more volatile implementation of corr_midplane

pyharm.ana.reductions.flatten_thphi(dump, var, at=5, sum=False)

Return a th-phi slice or sum of var. Note sums are not GR-aware!

Parameters:

• at – which rank in X1 to take data from.
• sum – whether to sum all ranks. Overrides ‘at’.

pyharm.ana.reductions.flatten_xy(dump, var, at=None, sum=False)

Return an X-Y slice or sum of var. Note sums are not GR-aware!

Parameters:

• at – which rank in X2 to take data from, default N2//2
• sum – whether to sum all ranks. Overrides ‘at’.

pyharm.ana.reductions.flatten_xz(dump, var, at=None, sum=False, half_cut=False)

Return an X-Z slice or sum of var, generally for use in making a plot. By default takes both the 0-degree (right side) and 180-degree (left side) slices, to make a full slice across the pole. Note sums are not GR-aware!

Parameters:

• at – which rank in X3 to take data from
• sum – whether to sum all ranks. Overrides ‘at’
• half_cut – just take the rank ‘at’, not its 180-degree opposite

pyharm.ana.reductions.get_i_slice(dump, r_min, r_max)

Calculate slice imin:imax for a given range in r

pyharm.ana.reductions.get_j_bounds(dump, th_min=1.0471975511965976, th_max=2.0943951023931953)

Calculate slice jmin:jmax for a given range in theta: returns tuple Defaults to EHT disk profile: pi/3 to 2pi/3

pyharm.ana.reductions.get_j_slice(dump, th_min=1.0471975511965976, th_max=2.0943951023931953)

Calculate slice jmin:jmax for a given range in theta: returns slice Defaults to EHT disk profile: pi/3 to 2pi/3

pyharm.ana.reductions.midplane_sum(dump, var, zones=2, **kwargs)

Average a few zones adjacent to midplane, then sum. Allows specifying an r_slice or i_slice within which to sum.

pyharm.ana.reductions.pspec(var, dt=5, window=0.33, half_overlap=False, bin='fib')

Power spectrum of a 1D timeseries, with various windows/binning. Currently only supports equally spaced times dt.

pyharm.ana.reductions.shell_avg(dump, var, **kwargs)

Average a variable over spherical shells. Returns a radial profile (array length N1) or single-shell average. See shell_sum for arguments.

pyharm.ana.reductions.shell_sum(dump, var, at_r=None, at_i=None, th_slice=None, j_slice=None, mask=None)

Sum a variable over spherical shells. Returns a radial profile (array length N1) or single-shell sum

Parameters:

• at_r – Single radius at which to sum (nearest-neighbor smaller zone is used)
• at_zone – Specific radial zone at which to sum, for compatibility
• th_slice – Tuple of minimum and maximum theta value to sum
• j_slice – Tuple of x2 indices instead of specifying theta. Overrides th_slice
• mask – array of 1/0 of the post-slice size of ‘var’, which is multiplied with the result

pyharm.ana.reductions.sphere_avg(dump, var, **kwargs)

Average everything within a sphere, semi-sphere or thick spherical shell. See sphere_sum for arguments.

pyharm.ana.reductions.sphere_sum(dump, var, r_slice=None, i_slice=None, th_slice=None, j_slice=None, mask=None)

Sum everything within a sphere, semi-sphere, or thick spherical shell. Extents are specified optionally in r or i, and th or j, with indices taking precedence Mask is multiplied at the end

pyharm.ana.reductions.theta_profile(dump, var, start, zones_to_av=1, use_gdet=True, fold=True)

Profile in theta by averaging over phi at a particular radius (or average among a few close radii). Note that this function returns an array of size N2 if fold==False, N2//2 if fold==True

Parameters:

• start – zone number to average at
• zones_to_av – number of zones to average, starting at ‘start’
• use_gdet – whether or not to make averaging GR-aware
• fold – whether to consider the system as symmetric about the midplane and average hemispheres

pyharm.ana.reductions.wrap(x)

Append the first rank in axis 1 again after the last rank in axis 1 (i.e. the angular index, after any reductions).

Useful when plotting zone-centered th or phi variables, as it allows shading of the region between ends of the array eliminating an ugly gap.