Sample

class cosmic.sample.sampler.independent.Sample

Bases: object

Methods Summary

binary_select(primary_mass[, binfrac_model])	Select which primary masses will have a companion using either a binary fraction specified by a float or a primary-mass dependent binary fraction following van Haaften et al.(2009) in appdx or Offner et al.(2023) in fig 1
sample_SFH([SF_start, SF_duration, met, size])	Sample an evolution time for each binary based on a user-specified time at the start of star formation and the duration of star formation.
sample_ecc(aRL_over_a[, ecc_model, size])	Sample the eccentricity according to a user specified model
sample_porb(mass1, mass2, rad1, rad2, porb_model)	Sample the orbital period according to the user-specified model
sample_primary([primary_model, size])	Sample the primary mass (always the most massive star) from a user-selected model
sample_secondary(primary_mass[, q_power_law])	Sample a secondary mass using draws from a uniform mass ratio distribution motivated by Mazeh et al. (1992) and Goldberg & Mazeh (1994).
set_kstar(mass)	Initialize stellar types according to BSE classification kstar=1 if M>=0.7 Msun; kstar=0 if M<0.7 Msun
set_reff(mass, metallicity[, zsun, SSEDict])	Better way to set the radii from BSE, by calling it directly

Methods Documentation

binary_select(primary_mass, binfrac_model=0.5, **kwargs)

Select which primary masses will have a companion using either a binary fraction specified by a float or a primary-mass dependent binary fraction following van Haaften et al.(2009) in appdx or Offner et al.(2023) in fig 1

Parameters:

primary_massarray

Mass that determines the binary fraction

binfrac_modelstr or float

vanHaaften - primary mass dependent and ONLY VALID up to 100 Msun

offner23 - primary mass dependent

float - fraction of binaries; 0.5 means 2 in 3 stars are a binary pair while 1 means every star is in a binary pair

Optional kwargs are defined in `get_independent_sampler`

Returns:

stars_in_binaryarray

primary masses that will have a binary companion

stars_in_singlearray

primary masses that will be single stars

binary_fractionarray

system-specific probability of being in a binary

binaryIdxarray

Idx of stars in binary

sample_SFH(SF_start=13700.0, SF_duration=0.0, met=0.02, size=None)

Sample an evolution time for each binary based on a user-specified time at the start of star formation and the duration of star formation. The default is a burst of star formation 13,700 Myr in the past.

Parameters:

SF_startfloat

Time in the past when star formation initiates in Myr

SF_durationfloat

Duration of constant star formation beginning from SF_Start in Myr

metfloat

metallicity of the population [Z_sun = 0.02] Default: 0.02

sizeint, optional

number of evolution times to sample NOTE: this is set in cosmic-pop call as Nstep

Returns:

tphysarray

array of evolution times of size=size

metallicityarray

array of metallicities

sample_ecc(aRL_over_a, ecc_model='sana12', size=None)

Sample the eccentricity according to a user specified model

Parameters:

ecc_modelstring

‘thermal’ samples from a thermal eccentricity distribution following Heggie (1975) ‘uniform’ samples from a uniform eccentricity distribution ‘sana12’ samples from the eccentricity distribution from Sana+2012 <https://ui.adsabs.harvard.edu/abs/2012Sci…337..444S/abstract>_ ‘circular’ assumes zero eccentricity for all systems DEFAULT = ‘sana12’

aRL_over_aratio of the minimum seperation (where RL overflow starts)

to the sampled semi-major axis. Use this to truncate the eccentricitiy

sizeint, optional

number of eccentricities to sample this is set in cosmic-pop call as Nstep

Returns:

eccarray

array of sampled eccentricities with size=size

sample_porb(mass1, mass2, rad1, rad2, porb_model, porb_max=None, size=None, **kwargs)

Sample the orbital period according to the user-specified model

Parameters:

mass1array

primary masses

mass2array

secondary masses

rad1array

radii of the primaries.

rad2array

radii of the secondaries

porb_modelstr or dict

selects which model to sample orbital periods, choices include: log_uniform : semi-major axis flat in log space from RRLO < 0.5 up to 1e5 Rsun according to Abt (1983) and consistent with Dominik+2012,2013 and then converted to orbital period in days using Kepler III sana12 : power law orbital period between 0.15 < log(P/day) < 5.5 following Sana+2012 <https://ui.adsabs.harvard.edu/abs/2012Sci…337..444S/abstract>_ renzo19 : power law orbital period for m1 > 15Msun binaries from Sana+2012 <https://ui.adsabs.harvard.edu/abs/2012Sci…337..444S/abstract>_ following the implementation of Renzo+2019 <https://ui.adsabs.harvard.edu/abs/2019A%26A…624A..66R/abstract>_ and flat in log otherwise raghavan10 : log normal orbital periods in days with mean_logP = 4.9 and sigma_logP = 2.3 between 0 < log10(P/day) < 9 following Raghavan+2010 <https://ui.adsabs.harvard.edu/abs/2010ApJS..190….1R/abstract>_ moe19 : log normal orbital periods in days with mean_logP = 4.9 and sigma_logP = 2.3 between 0 < log10(P/day) < 9 following Raghavan+2010 <https://ui.adsabs.harvard.edu/abs/2010ApJS..190….1R/abstract>_ but with different close binary fractions following Moe+2019 <https://ui.adsabs.harvard.edu/abs/2019ApJ…875…61M/abstract>_ martinez26 : piecewise model with a power law orbital period following Sana+2012 <https://ui.adsabs.harvard.edu/abs/2012Sci…337..444S/abstract>_ between 0.15 < log(P/day) < log(3000) for binaries with primaries large enough to undergo an FeCCSN and following Raghavan+2010 <https://ui.adsabs.harvard.edu/abs/2010ApJS..190….1R/abstract>_ with a log normal orbital period in days with mean_logP = 4.9 and sigma_logP = 2.3 between 0 < log10(P/day) < 9 for binaries with low mass primaries. Used in Martinez+2026 <https://ui.adsabs.harvard.edu/abs/2025arXiv251123285M/abstract>_. Assumes zsun = 0.02 for the metallicity dependence to predict outcomes correctly. martinez26_ecsn : same as martinez26 but with the Sana+2012 power law orbital period distribution for primaries massive enough to undergo an ECSN instead of an FeCCSN. Custom power law distribution defined with a dictionary with keys “min”, “max”, and “slope” (e.g. porb_model={“min”: 0.15, “max”: 0.55, “slope”: -0.55}) would reproduce the Sana+2012 distribution.

metfloat

metallicity of the population

Returns:

porbarray

orbital period with array size equalling array size of mass1 and mass2 in units of days

aRL_over_a: array

ratio of radius where RL overflow starts to the sampled seperation used to truncate the eccentricitiy distribution

sample_primary(primary_model='kroupa01', size=None, **kwargs)

Sample the primary mass (always the most massive star) from a user-selected model

kroupa93 follows Kroupa (1993), normalization comes from Hurley 2002 between 0.08 and 150 Msun salpter55 follows Salpeter (1955) between 0.08 and 150 Msun kroupa01 follows Kroupa (2001) <https://arxiv.org/abs/astro-ph/0009005> between 0.08 and 100 Msun

Parameters:

primary_modelstr, optional

model for mass distribution; choose from:

kroupa93 follows Kroupa (1993), normalization comes from

`Hurley 2002 <https://arxiv.org/abs/astro-ph/0201220>`_

valid for masses between 0.1 and 100 Msun

salpter55 follows

`Salpeter (1955) <http://adsabs.harvard.edu/abs/1955ApJ…121..161S>`_

valid for masses between 0.1 and 100 Msun

kroupa01 follows Kroupa (2001), normalization comes from

`Hurley 2002 <https://arxiv.org/abs/astro-ph/0009005>`_

valid for masses between 0.1 and 100 Msun

custom is a generic piecewise power law that takes in the power

law slopes and break points given in the optional input lists (alphas, mcuts)

default alphas and mcuts yield an IMF identical to kroupa01

Default kroupa01

sizeint, optional

number of initial primary masses to sample

NOTE: this is set in cosmic-pop call as Nstep

alphasarray, optional

absolute values of the power law slopes for primary_model = ‘custom’

Default [-1.3,-2.3,-2.3] (identical to slopes for primary_model = ‘kroupa01’)

mcutsarray, optional, units of Msun

break points separating the power law ‘pieces’ for primary_model = ‘custom’

Default [0.08,0.5,1.0,150.] (identical to breaks for primary_model = ‘kroupa01’)

Optional kwargs are defined in `get_independent_sampler`

Returns:

a_0array

Sampled primary masses

np.sum(a_0)float

Total amount of mass sampled

sample_secondary(primary_mass, q_power_law=0, **kwargs)

Sample a secondary mass using draws from a uniform mass ratio distribution motivated by Mazeh et al. (1992) and Goldberg & Mazeh (1994)

NOTE: the lower lim is set by either qmin or m2_min which are passed as kwargs

Parameters:

primary_massarray

sets the maximum secondary mass (for a maximum mass ratio of 1)

Optional kwargs are defined in `get_independent_sampler`

Returns:

secondary_massarray

sampled secondary masses with array size matching size of primary_mass

set_kstar(mass)

Initialize stellar types according to BSE classification kstar=1 if M>=0.7 Msun; kstar=0 if M<0.7 Msun

Parameters:

massarray

array of masses

Returns:

kstararray

array of initial stellar types

set_reff(mass, metallicity, zsun=0.02, SSEDict=None, **kwargs)

Better way to set the radii from BSE, by calling it directly

takes masses and metallicities, and returns the radii

Note that the BSE function is hard-coded to go through arrays of length 10^5. If your masses are more than that, you’ll need to divide it into chunks