Revision bcf992a2ca60f4dc2ad7b81321de176d74b80297 authored by Tim Haines on 23 October 2018, 23:21:25 UTC, committed by Tim Haines on 10 November 2018, 04:35:03 UTC
Change-Id: Ic48c4ac92b03138cfdce71d0792caedf15a97d76
1 parent 41969a5
imf.h
#ifndef IMF_HINCLUDED
#define IMF_HINCLUDED
// Please see
// http://charm.cs.uiuc.edu/manuals/html/charm++/3_17.html#SECTION000317400000000000000
// for details about PUPing child and parent classes.
// http://www.parashift.com/c++-faq-lite/abcs.html#faq-22.5
// for copy constructors in inherited abstract classes and a
// description of pure virtual functions
/**
* @brief Interface class for initial mass function.
*/
class IMF : public PUP::able {
public:
IMF() {};
/** @brief return stars per unit logarithmic mass
@param mass in solar masses.
*/
virtual double returnimf(double mass) const = 0;
/** @brief Charm++ requirement for passing polymorphic objects. */
PUPable_abstract(IMF);
/** @brief Charm++ migrate constructor */
IMF(CkMigrateMessage *m) : PUP::able(m) {}
/** @brief Charm++ Pack-UnPack method */
virtual void pup(PUP::er &p) = 0;
/** @brief Cumulative number of stars with mass greater than mass.
@param mass in solar masses */
virtual double CumNumber(double mass) = 0;
/** @brief Cumulative mass of stars with mass greater than mass.
@param mass in solar masses */
virtual double CumMass(double mass) = 0;
/** @brief copy IMF object */
virtual IMF* clone() const = 0;
~IMF() {};
};
/** @brief Implement Miller-Scalo IMF
Uses the 3 segment power law fit for the Miller-Scalo IMF
(Ap.J. Supp., 41,1979).
a1*(M**(b1)) 0.1<M<1.
IMF(Log10(M))= a2*(M**(b2)) 1.<M<10.
a3*(M**(b3)) 10.<M
Miller-Scalo IMF (Miller & Scalo, Ap.J. Supp., 41, 513, 1979) in
stars per unit logarithmic mass. Divide by M (mass) for IMF in
stars per unit mass. Also IMF is defined per yer per pc^2,
integrated over a cylinder that extends "several hundred parsecs on
either side of the plane of the galaxy"
*/
class MillerScalo : public IMF {
double a1, b1, m1;
double a2, b2, m2;
double a3, b3, m3;
double mmax;
public:
/* normalization, index, minimum mass */
MillerScalo() {
a1=42.0; b1=-0.4;m1=0.1; /* parameters from Ap.J. Supp., 41,1979 */
a2=42.0;b2=-1.5;m2=1.0; /* This is discontinuous, but is what */
a3=240.0;b3=-2.3;m3=10.0;/* they report in paper, so we leave it.*/
mmax=100.0;
}
/** @brief Charm++ method for migrating derived classes */
PUPable_decl(MillerScalo);
/** @brief Charm++ migration constructor */
MillerScalo(CkMigrateMessage *m) : IMF(m) {}
virtual double returnimf(double mass) const;
virtual double CumNumber(double mass);
virtual double CumMass(double mass);
virtual MillerScalo* clone() const;
virtual void pup(PUP::er &p) {
PUP::able::pup(p);
p|a1; p|b1; p|m1;
p|a2; p|b2; p|m2;
p|a3; p|b3; p|m3;
p|mmax;
}
};
/**
* @brief Implement IMF from Kroupa, Tout & Gilmore, 1993
*/
class Kroupa93 : public IMF {
double a1, b1, m1;
double a2, b2, m2;
double a3, b3, m3;
double mmax;
public:
/* parameters from Raiteri et. al. A&A, 315,1996, eq. 2; See also the
conclusions of Kroupa, Tout & Gilmore, 1993. */
/* To convert to the IMF(log10(M)) convention of Miller-Scale, we
increase the power law by 1 and multiply the coefficient by
ln(10.0). See, eg., Chabrier 2003, eq. 2 */
Kroupa93() {a1=0.3029*1.86606*log(10.0);b1=-0.3;m1=.08;
a2=0.3029*log(10.0);b2=-1.2;m2=0.5;
a3=0.3029*log(10.0); b3=-1.7; m3=1.0;
mmax=100.0; }
/** @brief Charm++ method for migrating derived classes */
PUPable_decl(Kroupa93);
/** @brief Charm++ migration constructor */
Kroupa93(CkMigrateMessage *m) : IMF(m) {}
virtual double returnimf(double mass) const;
virtual double CumNumber(double mass);
virtual double CumMass(double mass);
virtual Kroupa93* clone() const;
virtual void pup(PUP::er &p) {
PUP::able::pup(p);
p|a1; p|b1; p|m1;
p|a2; p|b2; p|m2;
p|a3; p|b3; p|m3;
p|mmax;
}
};
/**
* @brief Implement IMF from Kroupa 2001.
*/
class Kroupa01 : public IMF {
double a1, b1, m1;
double a2, b2, m2;
double mmax;
public:
/* parameters from Kroupa 2001, equation 2, and ignoring brown dwarfs,
Also normalized so that the mass integral is 1. */
/* NOTE BENE: Kroupa 2001 has a revised IMF in section 6.2 which is
different than this; however, below is what is used as the default in
Starburst99
(http://www.stsci.edu/science/starburst99/mappings/docs/run.html)
with the exception that the low mass cutoff is .1 instead of the .08
below and in the Kroupa paper.
*/
/* To convert to the IMF(log10(M)) convention of Miller-Scalo, we
increase the power law by 1 and multiply the coefficient by
ln(10.0). See, eg., Chabrier 2003, eq. 2 */
Kroupa01() {
a1=0.22038*2.0*log(10.0);b1=-0.3;m1=.08;
a2=0.22038*log(10.0);b2=-1.3;m2=0.5;
mmax=100.0; }
/** @brief Charm++ method for migrating derived classes */
PUPable_decl(Kroupa01);
/** @brief Charm++ migration constructor */
Kroupa01(CkMigrateMessage *m) : IMF(m) {}
virtual double returnimf(double mass) const;
virtual double CumNumber(double mass);
virtual double CumMass(double mass);
virtual Kroupa01* clone() const;
virtual void pup(PUP::er &p) {
PUP::able::pup(p);
p|a1; p|b1; p|m1;
p|a2; p|b2; p|m2;
p|mmax;
}
};
/** @brief Implement Chabrier IMF
Use the log normal + power law fit of Chabrier, 2003, Galactic
Stellar and Substellar Initial Mass Function", PASP 115, 763.
*/
class Chabrier : public IMF {
double a1, b1, m1;
double a2, b2, m2;
double mmax;
public:
/*
Chabrier low mass formula:
\xi(log m) = A exp [ - (log m - log m_c)^2/2 \sigma^2]
double a1, sigma, mc; (b1 is sigma, m1 is mc)
*/
Chabrier() {
a1=0.158;b1=0.69;m1=.079;
/* For high mass: normalization, index, minimum mass */
/* Parameters from Table 1 of Chabrier, 2003. */
a2=4.43e-2;b2=-1.3; m2=1.0;
mmax=100.0;
}
virtual double returnimf(double mass) const;
virtual double CumNumber(double mass);
virtual double CumMass(double mass);
virtual Chabrier* clone() const;
/** @brief Charm++ method for migrating derived classes */
PUPable_decl(Chabrier);
/** @brief Charm++ migration constructor */
Chabrier(CkMigrateMessage *m) : IMF(m) {}
virtual void pup(PUP::er &p) {
PUP::able::pup(p);
p|a1; p|b1; p|m1;
p|a2; p|b2; p|m2;
p|mmax;
}
};
#endif
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...