CAMBError: Error in Fortran called from calc_transfer: binary_search (e.g for optical depth) did not converge: Base range 800.00-2200.0.

[Eminencetoosure]

Creating the final list to be fed into the pipeline

result_list = idx.tolist()

final_list = list() for val in result_list:

run the pipeline for the given parameters. All results are saved in block

block    =  pipeline.run_parameters(val)

stats_name_cosmosis = "cosebis"     


# read in cosebis from block 
nbin_a=block["shear_cl","nbin_a"] 
nbin_b=block["shear_cl","nbin_b"] 
counter=0
for i_bin in range(1, nbin_a+1): 
    for j_bin in range(1, nbin_b+1): 
        name_in="bin_"+str(j_bin)+("_")+str(i_bin)
        if(block.has_value(stats_name_cosmosis, name_in)):
            th = block[stats_name_cosmosis,name_in]
            if(counter==0):
                th_all_3=th.copy() 
            else: 
                th_all_3=np.hstack((th_all_3,th))
                final_list.append(th_all_3)

final_list=np.hstack((th,th_all_3))

            counter+=1 

sample_S8 took: 0.000 seconds Warning: Integrate_Romberg failed to converge; integral, error, tol: 1.8779214533333051 -2.0412243158021148E-003 1.0000000000000001E-007 Warning: Integrate_Romberg failed to converge; integral, error, tol: 0.72499700489964747 -8.3126111458664553E-006 1.0000000000000001E-007 Warning: Integrate_Romberg failed to converge; integral, error, tol: 1.8779214533333051 -2.0412243158021148E-003 1.0000000000000001E-007 sigma8toAs took: 61.704 seconds Warning: Integrate_Romberg failed to converge; integral, error, tol: 1.8779214533333051 -2.0412243158021148E-003 1.0000000000000001E-007

CAMBError Traceback (most recent call last) Input In [40], in <cell line: 5>() 4 final_list = list() 5 for val in result_list: 6
7 # run the pipeline for the given parameters. All results are saved in block ----> 8 block = pipeline.run_parameters(val) 10 stats_name_cosmosis = "cosebis"
13 # read in cosebis from block

File ~/base/lib/python3.9/site-packages/cosmosis/runtime/pipeline.py:1110, in LikelihoodPipeline.run_parameters(self, p, check_ranges, all_params) 1097 u"""Assemble :class:DataBlock data based on parameter values in p, and run the pipeline on those data. 1098 1099 If check_ranges is indicated, the function will return None if (...) 1106 1107 """ 1108 data = self.build_starting_block(p, check_ranges=check_ranges, all_params=all_params) -> 1110 if self.run(data): 1111 # First run. If we have not set the likelihood names in the parameter 1112 # file then get them from the block 1113 if self.likelihood_names == NO_LIKELIHOOD_NAMES: 1114 self._set_likelihood_names_from_block(data)

File ~/base/lib/python3.9/site-packages/cosmosis/runtime/pipeline.py:631, in Pipeline.run(self, data_package) 628 if self.timing: 629 t1 = time.time() --> 631 status = module.execute(data_package) 633 if status is None: 634 raise ValueError(("A module you ran, '{}', did not return a proper status value.\n"+ 635 "It should return an integer, 0 if everything worked.\n"+ 636 "Sorry to be picky but this kind of thing is important.").format(module))

File ~/base/lib/python3.9/site-packages/cosmosis/runtime/module.py:249, in Module.execute(self, data_block) 247 raise RuntimeError("Must set up module before executing it") 248 if self.data is not None: --> 249 return self.execute_function(data_block, self.data) 250 else: 251 return self.execute_function(data_block)

File ~/Downloads//kcap//cosmosis-standard-library//boltzmann/pycamb/camb_interface.py:342, in execute(block, config) 340 r = camb.get_background(p) 341 else: --> 342 r = camb.get_results(p) 344 # Write derived parameters to cosmological_parameters section 345 derived = r.get_derived_params()

File ~/base/lib/python3.9/site-packages/camb/camb.py:35, in get_results(params) 33 if _debug_params: 34 print(params) ---> 35 res.calc_power_spectra(params) 36 return res

File ~/base/lib/python3.9/site-packages/camb/results.py:333, in CAMBdata.calc_power_spectra(self, params) 326 """ 327 Calculates transfer functions and power spectra. 328 329 :param params: optional :class:~.model.CAMBparams instance with parameters to use 330 331 """ 332 if params is not None: --> 333 self.calc_transfers(params, only_transfers=False) 334 else: 335 self._check_powers()

File ~/base/lib/python3.9/site-packages/camb/results.py:315, in CAMBdata.calc_transfers(self, params, only_transfers, only_time_sources) 312 self._check_powers(params) 313 if CAMBdata_gettransfers(byref(self), byref(params), byref(c_int(1 if only_transfers else 0)), 314 byref(c_int(1 if only_time_sources else 0))): --> 315 config.check_global_error('calc_transfer')

File ~/base/lib/python3.9/site-packages/camb/_config.py:51, in _config.check_global_error(self, reference) 49 reference = '' 50 if err: ---> 51 raise CAMBError(reference + '%s' % err) 52 else: 53 raise CAMBError(reference + 'Error code: %s' % code)

CAMBError: Error in Fortran called from calc_transfer: binary_search (e.g for optical depth) did not converge: Base range 800.00-2200.0.

The above error was gotten when i was trying to pass my LHS sample of data to my cosmosis pipeline, can you please help me to solve this problem

[cmbant]

What were the parameters? Perhaps you forgot to fix tau?

[Eminencetoosure]

import numpy as np import pyDOE as pyDOE

number of parameters and samples

n_params = 7 n_samples = 400000

parameter ranges

obh2 = np.linspace(0.019, 0.026, n_samples) omch2 = np.linspace(0.051, 0.255, n_samples) h0 = np.linspace(0.64, 0.82, n_samples) n_s = np.linspace(0.84, 1.1, n_samples) s_8_input = np.linspace(0.1, 1.3, n_samples) logt_agn = np.linspace(7.6, 8.0, n_samples) A = np.linspace(-6.0, 6.0, n_samples)

LHS grid

AllParams = np.vstack([obh2, omch2, h0, n_s, s_8_input, logt_agn, A]) lhd = pyDOE.lhs(n_params, samples=n_samples, criterion=None) idx = (lhd * n_samples).astype(int) print(type(idx)) print(len(idx))

The above is the range of my cosmosis parameter. i used this code to create LHS, and i pass the idx(which is the LHS sample) to my cosmosis pipeline to generate the theoretical_value(th_all_3). th_all_3 is supposed to be my cosmosis dataset which im going to train using cosmopower but i for the error ...

Please how can you help me fix this err

[cmbant]

You need to look at the parameters being sent to CAMB, I can't help with Cosmosis. For a CAMB bug report, sample code just calling CAMB directly is really needed.