-
Notifications
You must be signed in to change notification settings - Fork 26
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
3 changed files
with
159 additions
and
31 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -0,0 +1,115 @@ | ||
| import numpy as np | ||
| import healpy as hp | ||
| import pymaster as nmt | ||
|
|
||
|
|
||
| def test_anisotropic_weighting(): | ||
| # Test parameters | ||
| nside = 64 | ||
| spin = 2 | ||
| nlb = 4 | ||
| nsims = 100 | ||
|
|
||
| # Create anisotropic weights | ||
| mask = hp.read_map("test/benchmarks/msk.fits") | ||
| mask = hp.ud_grade(mask, nside_out=nside) | ||
| mask = hp.smoothing(mask, sigma=np.radians(1.0)) | ||
| mask[mask > 1] = 1 | ||
| mask[mask < 0.001] = 0 | ||
|
|
||
| delta_m = 0.9 | ||
| r_m = 0.5 | ||
| w11 = (1+delta_m)*mask | ||
| w22 = (1-delta_m)*mask | ||
| w12 = r_m*np.sqrt(w11*w22) | ||
|
|
||
| # Input power spectra | ||
| ls = np.arange(3*nside) | ||
| cl_temp = 1/(ls+10) | ||
| cl_tt = 1.5*cl_temp | ||
| cl_te = 0.6*cl_temp | ||
| cl_tb = 0.3*cl_temp | ||
| cl_ee = 1.0*cl_temp | ||
| cl_eb = 0.2*cl_temp | ||
| cl_bb = 0.4*cl_temp | ||
|
|
||
| # Workspaces | ||
| b = nmt.NmtBin.from_nside_linear(nside, nlb=nlb) | ||
| leff = b.get_effective_ells() | ||
| f0 = nmt.NmtField(mask, None, spin=0, n_iter=0) | ||
| fs = nmt.NmtField(mask, None, spin=spin, n_iter=0) | ||
| fsa = nmt.NmtField(w11, None, spin=spin, n_iter=0, | ||
| mask_12=w12, mask_22=w22) | ||
| w0sa = nmt.NmtWorkspace.from_fields(f0, fsa, b) | ||
| wsa0 = nmt.NmtWorkspace.from_fields(fsa, f0, b) | ||
| wsasa = nmt.NmtWorkspace.from_fields(fsa, fsa, b) | ||
| wssa = nmt.NmtWorkspace.from_fields(fs, fsa, b) | ||
| wsas = nmt.NmtWorkspace.from_fields(fsa, fs, b) | ||
|
|
||
| # Run simulations | ||
| cl0sa_s = [] | ||
| clsa0_s = [] | ||
| clsasa_s = [] | ||
| clssa_s = [] | ||
| clsas_s = [] | ||
| for i in range(nsims): | ||
| almt, alme, almb = hp.synalm([cl_tt, cl_ee, cl_bb, | ||
| cl_te, cl_eb, cl_tb], new=True) | ||
| map_t = hp.alm2map(almt, nside, lmax=3*nside-1) | ||
| map_q, map_u = hp.alm2map_spin([alme, almb], nside, spin, | ||
| lmax=3*nside-1, mmax=3*nside-1) | ||
|
|
||
| f0 = nmt.NmtField(mask, [map_t], n_iter=0) | ||
| fs = nmt.NmtField(mask, [map_q, map_u], spin=spin, n_iter=0) | ||
| fsa = nmt.NmtField(w11, [map_q, map_u], spin=spin, n_iter=0, | ||
| mask_12=w12, mask_22=w22) | ||
|
|
||
| cl0sa_s.append(w0sa.decouple_cell(nmt.compute_coupled_cell(f0, fsa))) | ||
| clsa0_s.append(wsa0.decouple_cell(nmt.compute_coupled_cell(fsa, f0))) | ||
| clsasa_s.append(wsasa.decouple_cell(nmt.compute_coupled_cell(fsa, | ||
| fsa))) | ||
| clssa_s.append(wssa.decouple_cell(nmt.compute_coupled_cell(fs, fsa))) | ||
| clsas_s.append(wsas.decouple_cell(nmt.compute_coupled_cell(fsa, fs))) | ||
| cl0sa_s = np.array(cl0sa_s) | ||
| clsa0_s = np.array(clsa0_s) | ||
| clsasa_s = np.array(clsasa_s) | ||
| clssa_s = np.array(clssa_s) | ||
| clsas_s = np.array(clsas_s) | ||
| # Mean | ||
| cl0sa_m = np.mean(cl0sa_s, axis=0) | ||
| clsa0_m = np.mean(clsa0_s, axis=0) | ||
| clsasa_m = np.mean(clsasa_s, axis=0) | ||
| clssa_m = np.mean(clssa_s, axis=0) | ||
| clsas_m = np.mean(clsas_s, axis=0) | ||
| # STD | ||
| cl0sa_e = np.std(cl0sa_s, axis=0) | ||
| clsa0_e = np.std(clsa0_s, axis=0) | ||
| clsasa_e = np.std(clsasa_s, axis=0) | ||
| clssa_e = np.std(clssa_s, axis=0) | ||
| clsas_e = np.std(clsas_s, axis=0) | ||
| # Truth | ||
| cl0sa_t = w0sa.decouple_cell(w0sa.couple_cell([cl_te, cl_tb])) | ||
| clsa0_t = wsa0.decouple_cell(wsa0.couple_cell([cl_te, cl_tb])) | ||
| clsasa_t = wsasa.decouple_cell(wsasa.couple_cell([cl_ee, cl_eb, | ||
| cl_eb, cl_bb])) | ||
| clssa_t = wssa.decouple_cell(wssa.couple_cell([cl_ee, cl_eb, | ||
| cl_eb, cl_bb])) | ||
| clsas_t = wsas.decouple_cell(wsas.couple_cell([cl_ee, cl_eb, | ||
| cl_eb, cl_bb])) | ||
|
|
||
| # Compare all power spectra and check for > 6sigma deviations | ||
| def comp_cl(clm, cle, clt, nsigma=6): | ||
| lgood = leff < 2*nside | ||
| for m, e, t in zip(clm, cle, clt): | ||
| r = ((m-t)*np.sqrt(nsims)/e)[lgood] | ||
| if np.any(np.fabs(r) > nsigma): | ||
| return False | ||
| return True | ||
|
|
||
| test_0sa = comp_cl(cl0sa_m, cl0sa_e, cl0sa_t) | ||
| test_sa0 = comp_cl(clsa0_m, clsa0_e, clsa0_t) | ||
| test_sasa = comp_cl(clsasa_m, clsasa_e, clsasa_t) | ||
| test_ssa = comp_cl(clssa_m, clssa_e, clssa_t) | ||
| test_sas = comp_cl(clsas_m, clsas_e, clsas_t) | ||
|
|
||
| assert np.all([test_0sa, test_sa0, test_sasa, test_ssa, test_sas]) |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters