Class Documentation

bamr_class

class bamr_class

Compute neutron star structure for each MCMC point.

There will be a number of instances of this class equal to the number of OpenMP threads, all of which share the same neutron star data (in nsd) and settings (in set). They will each have their own model object, stored in mod, which is set in mcmc_bamr::set_model() .

Public Functions

bamr_class()

virtual ~bamr_class()

virtual int compute_point(const ubvector &pars, std::ofstream &scr_out, double &weight, model_data &dat)

Compute the EOS corresponding to parameters in e and put output in tab_eos.

virtual int fill(const ubvector &pars, double weight, std::vector<double> &line, model_data &dat)

Fill vector in line with data from the current Monte Carlo point.

Public Members

double schwarz_km

The Schwarzchild radius in km.

std::shared_ptr<ns_data> nsd

Pointer to neutron star data.

std::shared_ptr<const settings> set

Pointer to settings object.

std::shared_ptr<model> mod

Model object.

std::string model_type

Model type string.

mcmc_bamr

class mcmc_bamr : public o2scl::mcmc_para_cli<point_funct, fill_funct, model_data, ubvector>

Statistical analysis of EOS from M and R constraints.

Main functions called from the command-line interface

virtual int set_model(std::vector<std::string> &sv, bool itive_com)

Set the model for the EOS to use.

Public Functions

mcmc_bamr()

Create a mcmc_bamr object.

virtual ~mcmc_bamr()

virtual void setup_cli()

Set up the ‘cli’ object.

This function adds three commands (mcmc, model, add-data) and the ‘set’ parameters.

Public Members

o2scl::cli cl

The command-line interface object.

Protected Functions

virtual void file_header(o2scl_hdf::hdf_file &hf)

Write initial data to HDF file.

virtual int mcmc_init()

Make any necessary preparations for the mcmc() function.

This is called by mcmc(). If the return value is non-zero then it is assumed that the calculation fails and mcmc() returns.

virtual int add_data(std::vector<std::string> &sv, bool itive_com)

Add a data distribution to the list.

virtual int add_data_alt(std::vector<std::string> &sv, bool itive_com)

Add a data distribution to the list.

virtual int mcmc_func(std::vector<std::string> &sv, bool itive_com)

Perform the MCMC simulation.

virtual int threads(std::vector<std::string> &sv, bool itive_com)

Set the number of OpenMP threads.

virtual int initial_point_last(std::vector<std::string> &sv, bool itive_com)

Use the last point in a specifed file for the initial point.

virtual int initial_point_best(std::vector<std::string> &sv, bool itive_com)

Use the highest likelihood point in the specified file for the initial point.

virtual int read_prev_results(std::vector<std::string> &sv, bool itive_com)

Read previous results from a file.

Protected Attributes

std::string model_type

A string indicating which model is used, set in set_model().

std::vector<bamr_class *> bc_arr

Vector of bamr_class objects (one for each OpenMP thread)

This is currently a pointer because it makes it a lot easier to replace these pointers with children. A shared_ptr might be better, but I’ve had problems implementing vector<shared_ptr> correctly.

std::shared_ptr<settings> set

The bamr::settings object (shared by instances of bamr_class)

std::shared_ptr<ns_data> nsd

The bamr::ns_data object (shared by instances of bamr_class)

model_data

class model_data

Data at each MC point.

Public Functions

model_data()

model_data(const model_data &md)

Copy constructor.

model_data &operator=(const model_data &md)

Make operator= copy constructor private.

Public Members

ubvector rad

Radii.

ubvector mass

Masses.

ubvector wgts

Weights.

o2scl::table_units mvsr

M vs. R data.

o2scl::table_units eos

EOS data.

settings

class settings

Settings object.

Parameter objects for the ‘set’ command

o2scl::cli::parameter_double p_min_max_mass

o2scl::cli::parameter_double p_exit_mass

o2scl::cli::parameter_double p_input_dist_thresh

o2scl::cli::parameter_double p_min_mass

o2scl::cli::parameter_int p_grid_size

o2scl::cli::parameter_int p_mass_switch

o2scl::cli::parameter_int p_verbose

o2scl::cli::parameter_bool p_debug_star

o2scl::cli::parameter_bool p_debug_load

o2scl::cli::parameter_bool p_debug_eos

o2scl::cli::parameter_bool p_baryon_density

o2scl::cli::parameter_bool p_use_crust

o2scl::cli::parameter_bool p_inc_baryon_mass

o2scl::cli::parameter_bool p_norm_max

o2scl::cli::parameter_bool p_compute_cthick

o2scl::cli::parameter_bool p_addl_quants

o2scl::cli::parameter_bool p_crust_from_L

o2scl::cli::parameter_bool p_mpi_load_debug

o2scl::cli::parameter_double p_nb_low

o2scl::cli::parameter_double p_nb_high

o2scl::cli::parameter_double p_e_low

o2scl::cli::parameter_double p_e_high

o2scl::cli::parameter_double p_m_low

o2scl::cli::parameter_double p_m_high

o2scl::cli::parameter_double p_mvsr_pr_inc

Limits on mass and radius from source data files

These are automatically computed in load_mc() as the smallest rectangle in the \( (M,R) \) plane which encloses all of the user-specified source data

double in_m_min

double in_m_max

double in_r_min

double in_r_max

Other parameters accessed by ‘set’ and ‘get’

int grid_size

Number of bins for all histograms (default 100)

double mvsr_pr_inc

Pressure increment for the M vs. R curve (default 1.1)

bool norm_max

If true, normalize the data distributions so that the max is one, otherwise, normalize so that the integral is one (default false)

bool use_crust

If true, use the default crust (default true)

bool debug_load

If true, output debug information about the input data files (default false)

bool debug_star

If true, output stellar properties for debugging (default false)

bool debug_eos

If true, output equation of state for debugging (default false)

bool baryon_density

If true, compute the baryon density (default true)

double input_dist_thresh

The lower threshold for the input distributions (default 0.0)

double exit_mass

The upper mass threshold (default 10.0)

bool mpi_load_debug

If true, debug MPI file loading.

double min_mass

Minimum mass allowed for any of the individual neutron stars (default 1.0)

double min_max_mass

Minimum allowed maximum mass (default 2.0)

bool best_detail

If true, output more detailed information about the best point (default false)

bool inc_baryon_mass

If true, output information about the baryon mass as well as the gravitational mass (default false)

bool compute_cthick

If true, compute crust thicknesses (default false)

bool addl_quants

If true (default false)

bool crust_from_L

If true, compute a crust consistent with current value of L.

Only works if use_crust, baryon_density, and compute_cthick are true and the model provides S and L.

int mass_switch

Histogram limits

double nb_low

double nb_high

double e_low

double e_high

double m_low

double m_high

Public Functions

settings()

void setup_cli(o2scl::cli &cl)

Add parameters to the o2scl::cli object.

Public Members

int verbose

Verbosity parameter.

ns_data

class ns_data

Neutron star data object.

This class is designed so that multiple OpenMP threads can use the same const instance of this class (i.e. so long as they do not change the member data).

Input neutron star data

std::vector<o2scl::table3d> source_tables

Input probability distributions.

std::vector<o2scl::table3d> source_tables_alt

Alternate input probability distributions.

std::vector<std::string> source_names

The names for each source.

std::vector<std::string> table_names

The names of the table in the data file.

std::vector<std::string> source_fnames

File names for each source.

std::vector<std::string> source_fnames_alt

Alternate file names for each source.

std::vector<std::string> slice_names

Slice names for each source.

std::vector<double> init_mass_fracs

The initial set of neutron star masses.

size_t n_sources

The number of sources.

virtual int add_data(std::vector<std::string> &sv, bool itive_com)

Add a data distribution to the list.

virtual int add_data_alt(std::vector<std::string> &sv, bool itive_com)

Add a data distribution to the list.

virtual void load_mc(std::ofstream &scr_out, int mpi_nprocs, int mpi_rank, std::shared_ptr<settings> set)

Load input probability distributions.

Ensure all MPI ranks read all files while ensuring that no two ranks simultaneously read the same file.

Let MPI_Size be denoted n. If n is 1, no messages are sent and the single rank always reads the files in order. When n is larger than 1, rank 0 still reads the files in order. Rank n-1 begins by reading the second file, rank n-2 begins by reading the third file, and so on, until there are no more ranks left which are not reading files or there are no more files left. After reading, all ranks which read files send a message to the next rank (in sequence) prompting them to proceed with the next file. Once each rank has read all of the files, it does not send or receive any further messages (the number of messages sent for n>1 is n times the number of files times). If settings::mpi_load_debug is set to 1, then the various MPI messages are copied to scr_out, no files are actually read, and exit() is called after all messages have been sent and received by all ranks.

Public Functions

ns_data()

nstar_cold2

class nstar_cold2 : public o2scl::nstar_cold

A simplified version of nstar_cold.

This simplified version only computes the energy density and pressure rather than the version which computes several extra quantities. This class is part of bamr::model.

Public Functions

int calc_eos(double np_0 = 0.0)

Compute the core EOS.

process

class process

Process MCMC data from bamr

Command-line parameter objects

o2scl::cli::parameter_double p_xscale

o2scl::cli::parameter_double p_yscale

o2scl::cli::parameter_bool p_errors

o2scl::cli::parameter_bool p_logx

o2scl::cli::parameter_bool p_logy

o2scl::cli::parameter_bool p_logz

o2scl::cli::parameter_int p_hist_size

o2scl::cli::parameter_int p_n_blocks

o2scl::cli::parameter_int p_line_start

o2scl::cli::parameter_int p_verbose

o2scl::cli::parameter_string p_constraint

o2scl::cli::parameter_string p_weights

Command-line parameters

int verbose

Verbosity (default 1)

int line_start

Ignore all lines before start.

double xscale

Scale for x-axis.

double yscale

Scale for y-axis.

bool logx

If true, use a logarithmic x scale.

bool logy

If true, use a logarithmic y scale.

bool logz

If true, use a logarithmic z scale.

bool errors

If true, plot errors in 1d.

int hist_size_int

Histogram size.

int n_blocks

Number of blocks.

std::string weights_col

Column representing weights.

Axis limit variables

bool xset

If true, x limits are set.

double user_xlow

Lower x value.

double user_xhigh

Upper x value.

bool yset

If true, y limits are set.

double user_ylow

Lower y value.

double user_yhigh

Upper y value.

Const confidence limits

const double one_sigma

The one-sigma limit of a normal distribution, 0.6827.

const double two_sigma

The two-sigma limit of a normal distribution, 0.9545.

const double three_sigma

The three-sigma limit of a normal distribution, 0.9973.

Other class member data

std::string constraint

Constraint to apply to the data.

std::vector<double> cont_levels

Contour levels set in contours.

o2scl::format_float ff

Formatter for floating point numbers.

By default, this is set in the constructor to

ff.latex_mode();
ff.set_sig_figs(4);
ff.set_pad_zeros(true);
ff.set_exp_limits(-5,5);

ubvector scale

Scaling for each element in Bayes factor distance metric.

std::vector<std::vector<double>> bfactor_data

Storage for Bayes factor computation.

Commands

int auto_corr(std::vector<std::string> &sv, bool itive_com)

Create a table of autocorrelation data from a specified column.

int xlimits(std::vector<std::string> &sv, bool itive_com)

Set limits for the x-axis.

int ylimits(std::vector<std::string> &sv, bool itive_com)

Set limits for the y-axis.

int hist(std::vector<std::string> &sv, bool itive_com)

Create a histogram from a specified column.

int hist2(std::vector<std::string> &sv, bool itive_com)

Create a two-dimensional histogram from two user-specified columns.

int hist_set(std::vector<std::string> &sv, bool itive_com)

Create a set of histograms from a set of columns in the bamr MCMC output.

int curve_set(std::vector<std::string> &sv, bool itive_com)

Desc.

int combine(std::vector<std::string> &sv, bool itive_com)

Combine several bamr output files.

int contours(std::vector<std::string> &sv, bool itive_com)

Specify which contour levels to use.

int mass_sel(std::vector<std::string> &sv, bool itive_com)

Create new columns at a selected mass.

Internal functions

void swap(double &w1, size_t &i1, ubvector &x1, double &w2, size_t &i2, ubvector &x2)

Swap w1 and w2, i1 and i2, and x1 and x2 (for bfactor())

double dist_sq(const ubvector &x, size_t row)

Squared distance (for bfactor())

double func(size_t n, const ubvector &x)

Function to integrate for Bayes factor.

Public Functions

process()

Create the process object.

void run(int argc, char *argv[])

Main public interface.

Protected Functions

virtual void setup_cli()

Setup the command-line interface.

int bfactor(std::vector<std::string> &sv, bool itive_com)

Compute the Bayes factor.

int set_params_limits(std::vector<std::string> &sv, bool itive_com)

Set parameter names and upper and lower limits.

Protected Attributes

o2scl::cli cl

Command-line interface object.

std::vector<std::string> x_params

ubvector x_low

ubvector x_high