Tags: abacusorg/class_public
Tags
removed useless .ini and .pre files; added input files matching Planc… …k baseline models; struct_cleanup() now done automatically; several minor bug fixes * 2.7: updated version number to 2.7.1 and doc small improvements in setup.py with absolute directory paths completed the fix for minor precision issue, relevant only with k_output_values + rkck evolver fixed small things for a neat compilation even with -Wall Completed changes in output for _cb with isocurvature; output indented; not yet tested Fixed comments and indentation in _cb part of spectra.c P_cb changes reviewed and checked up to transfer.c included added comment and improved style of P_cb correction in nonlinear m odule. Yet to be tested. fixed minor precision issue, relevant only with k_output_values + rkck evolver Fixed small scanf counting mistake. updated the set of .pre files updated the set of .ini files Fixed whitespace and indentation. fixed a problem in the manual from release 2.7, and removed unnecessary files from the RSI Python wrapper: Prevent memory leaks from occuring. fixed small inconsistency in many_times.ipynb leading to errors with linux Added handling of printouts, which were generating spurious warnings during compilation. Furthermore, the code should be slightly safer now.
Main new features are P_cb, P_k_eq, RealSpaceInterface, new notebooks… … and few bugfixes * devel: (54 commits) updated doc added a few comments in spectra related to P_cb updated version number to 2.7.0 fixed minor issue with dcdm in input restored the empty directory RealSpaceInterface/cache/ added 11 notebooks and equivalent scripts (cambridge version) pushing the Real Space Interface further cosmetic changes in P_cb few extra comkments to Pcb part Fixed a bug leading to double free when combining the new quadrature strategies with shooting Fixed typo in ncdm number density, cf. issue lesgourg#225 final polishing of pk_eq method taken care of the case thatan x-array with 2 entries only needs to be splined by using the natural_spline and not the estimated derivative spline method by default in this particular case sigma8 and sigma8_cb now computed in two separate functions to avoid confusion mem leak fix k_nl Pcb changes: M and CB share the same tau_min_nl Nils pk_type trick transfer.c dNdz according to Euclid IST specification removed the enum halofit_found_k_max ...
Few bug fixes and mnior improvements, especially on Halofit precision * 2.6: changed version number to 2.6.2 add sigma8 as an input parameter using shooting method with target A_s Added the new quadrature functionality Added check for mvec library in glibc to avoid issue lesgourg#99 changed default value of k_max_for_pk to 1/Mpc for precise sigma8 even with default settings fixed new precision settings for halofit just debugged previous commit changes in spectra module: store pk_nl in table with size independent from pk_l, for more correct interpolation improved coding style inside nonlinear_halofit added tau_reio() in classy now possible to ask for tensors only improved estimate of diffusion damping scale using R corrections from W. Hu added Omega_g() to classy (to get Omega_gamma today) corrected new BBN file: uses Yp, not Yp^BBN; added marcucci et al. version as an alternative corrected new BBN file: uses Yp, not Yp^BBN; added marcucci et al. version as an alternative updated BBN file
several bug and memory leak fixes * 2.6: changed version number to 2.6.1 corrected default value of selection_magnification_bias=0 corrected a factor in the definition of C_l^shearshear removed a memory leak in functions raw_cl ans lensed_cl removed a memory leak in functions pk and pk_lin of the wrapper added to the wrapper new functions f() and sigma() corrected bugs affecting only P_NL(k) in presence of isocurvature modes
fluid perturbations obey PPF approximation, w_fld(a) only in one plac… …e, python3 compatibility, new features for linear growth factor and transfer function output * devel: (21 commits) finsihed updates to doc for v2.6.0 updated first chapter of manual updated soxygen style, new background function compatible with doxygen better documented fluid parameters in explanatory.ini version number updated to 2.6.0 (but doc not yet updated) added details on w(a) in explanatory.ini done with w(a) rewriting w_fld(a) properly implemented everywhere else excepted 1 place in perturbations (w_prime) w_fld(a) properly implemented in background excepted 1 place (integral for IC) first working step in rewriting of w_fld(a) Fixed Makefile to not link classy with OpenMP if CLASS was compiled without OpenMP added output of Cl_dl in wrapper, following pull request lesgourg#23 (Credits Atsushi Nishizawa) new input flag: extra metric transfer functions, implementing pull request lesgourg#107 by Wessel Valkenburg precision parameter hyper_sampling_curved_low_nu changed from 6 to 7 new functions in classy for compatibility with SDSS likelihood in MontePython v2.3 added function scale_independent_growth_factor(z) to classy added reminders of lensing potential/deflection/shear conversion in explanatory and header Corrected D(a) to make it compatible with relativistic backscaling 1702.03221 Fixed a few Python3 compatibility issues and made the matter power spectrum computed check in the pk function simpler and more robust Implemented and tested non-zero curvature in PPF equations ...
Compatible with python3 (credits M. Millea), fixes for more robust wr… …apper installation, new reionisation scheme reio_inter * 2.5: new version number 2.5.1 another cosmetic change cosmetic changes in view of minor release more robust commands in Makefile for the python wrapper installation Makefile set back to coding system utf-8-unix to avoid problems of invisible characters adding PPF debug notebook started to code PPF in newtonian gauge reio_inter scheme: updated explanatory.ini added scheme reio_inter, but more checkes and debugging needed commented out tests on H0, T_cmb, Omega_k fixed other python 2/3 compatibility issues by hand ran 2to3 tool
new version with automatic documentation, curved galaxy number count,… … ceff2_ur, cvis2_ur, reio_many_tanh, new z_max_pk scheme, inflation_behavior = analytical * devel: (74 commits) Fixed a typo in the README Recompiled final version of documentation changes in chap3.md updated the end of Input section in chap3.md updated pk_ref.pre updated the README updated version numb er to v2.5.0 added small comments related to previous commits similar fix for case with sCl only fixed bug that prevented user form using nCl alone since commit 4c8c09a changes by Francesco concerning nCls Limber threshold and Bessel sampling added warning when k_NL of halofit too large; small bug fixes from Guido corrected small bug in spectra (relevant when outputing transfer function of dr in absnece of ur) corrected small error in recfast implementation, leading to negligible (~10^-6) error on Cls (Credits Sandro Pinto) corrected f_evo term in terms d0,d1 of number count Small formatting tweaks for the documentation completed README of doc Final version of the documentation changed a detail in a comment chapters 2 and 3 of documentation expanded ...
Corrected a few bugs affecting Halofit precision with massive nu or d… …ecaying DM * 2.4: version number changed to v2.4.5 included dcdm in definition of fnu in halofit series of fixes to remove numerical errors from halofit implementation relevant only for quintessence: corrected an abs() into fabs() renoved selection bias from shear spectrum, only relevant for number count (thanks Aurelien Benoit-Levy) corrected bug in interface to halofit, problematic only with decaying CDM (Omega_drdcdm>0) improved reco-reio matching, helps to get more precise z_reio(tau_reio) search Improved error handling for shooting method not to generate input errors for MontePython
PreviousNext