Example #1
0
# Left (vapor)
a_vap = 1e-10 / (RG.sigma ** 3 * np.pi / 6)
c_vap = nparticular
# Right (liquid)
a_liq = nparticular
c_liq = 0.55 / (RG.sigma ** 3 * np.pi / 6)
###############################

# Open file for output
fout = open("figs/npart_RG-out.dat", "w")

# label the columns of the output
fout.write("#T     nvapor     nliquid       phi(nvap)        phi(nliq)         nparticular\n")  #       phi_avg')

# Do first temperature before the loop
nvapor, phi_vapor = minmax_RG.minimize(RG.phi, T, a_vap, c_vap, nparticular, i)
print "nvap,phi_vap", nvapor, phi_vapor
nliquid, phi_liquid = minmax_RG.minimize(RG.phi, T, a_liq, c_liq, nparticular, i)
print "nliq,phi_liq", nliquid, phi_liquid
sys.stdout.flush()


# while T < 1.4:
for i in xrange(0, N + 1):
    T = (Tc - Tlow) * (1 - ((N - i) / N) ** 4) + Tlow

    fout.flush()
    # Starting point for new nparticular is abscissa of max RG.phi with old nparticular
    nparticular = minmax_RG.maximize(RG.phi, T, nvapor, nliquid, nparticular, i)

    # I'm looking at the minima of RG.phi
Example #2
0
def npart(iterations):
    # Initial conditions; dependent on you system

    T = 0.1
    nparticular = 0.08/(RG.sigma**3*np.pi/6) # using finite differences df_dn
    # nparticular = 0.155/(RG.sigma**3*np.pi/6) # using analytic df_nd

    # T = 0.68
    # nparticular = 0.0414/(RG.sigma**3*np.pi/6)

    N = 20 # For publication plots, make this bigger (40? 80? 100? You decide!)
    Tc = 1.33
    Tlow = T

    # Bounds of minimization.
    # Use range that works for many temperatures
    # Left (vapor)
    a_vap = 1e-10/(RG.sigma**3*np.pi/6)
    c_vap = nparticular
    # Right (liquid)
    a_liq = nparticular
    c_liq = 0.75/(RG.sigma**3*np.pi/6)
    ###############################

    # Open file for output
    fout = open('figs/npart_RG-i%d-out.dat'%iterations,'w')

    # label the columns of the output
    fout.write('#T     nvapor     nliquid       phi(nvap)        phi(nliq)         nparticular\n')#       phi_avg')
    sys.stdout.flush()

    # Do first temperature before the loop
    nvapor,phi_vapor = minmax_RG.minimize(RG.phi,T,a_vap,c_vap,nparticular,iterations)
    print 'initial nvap,phi_vap',nvapor,phi_vapor
    nliquid,phi_liquid = minmax_RG.minimize(RG.phi,T,a_liq,c_liq,nparticular,iterations)
    print 'initial nliq,phi_liq',nliquid,phi_liquid
    print 'initial npart',nparticular*(RG.sigma**3*np.pi/6)

    #while T < 1.4:
    for j in xrange(0,N+1):
        T = (Tc - Tlow)*(1 - ((N-j)/N)**4) + Tlow

        fout.flush()
        # Starting point for new nparticular is abscissa of max RG.phi with old nparticular
        nparticular = minmax_RG.maximize(RG.phi,T,nvapor, nliquid, nparticular,iterations)

        # I'm looking at the minima of RG.phi
        c_vap = nparticular
        a_liq = nparticular

        nvapor,phi_vapor = minmax_RG.minimize(RG.phi,T,c_vap,a_vap,nparticular,iterations)
        nliquid,phi_liquid = minmax_RG.minimize(RG.phi,T,a_liq,c_liq,nparticular,iterations)

        tol = 1e-5
        fnpart = RG.phi(T, nparticular, nparticular, iterations)

        # Compare the two minima in RG.phi
        while np.fabs(phi_vapor - phi_liquid)/np.fabs(fnpart) > tol:

            delta_mu = (phi_liquid - phi_vapor)/(nliquid - nvapor)

            def newphi(T, n, npart, i):
                return RG.phi(T, n, npart, i) - delta_mu*n

            nparticular = minmax_RG.maximize(newphi,T,nvapor,nliquid,nparticular,iterations)

            fnpart = RG.phi(T, nparticular, nparticular, iterations)

            nvapor,phi_vapor = minmax_RG.minimize(RG.phi,T,a_vap,c_vap,nparticular,iterations)
            nliquid,phi_liquid = minmax_RG.minimize(RG.phi,T,a_liq,c_liq,nparticular,iterations)

        fout.write(str(T))
        fout.write('  ')
        fout.write(str(nvapor))
        fout.write('  ')
        fout.write(str(nliquid))
        fout.write('  ')
        fout.write(str(phi_vapor))
        fout.write('  ')
        fout.write(str(phi_liquid))
        fout.write('  ')
        fout.write(str(nparticular))
        fout.write('\n')
        sys.stdout.flush();
        print '   T, etaVap, etaLiq',T,nvapor/(RG.sigma**3*np.pi/6),nliquid/(RG.sigma**3*np.pi/6)
    fout.close()
Example #3
0
def npart(iterations):
    # Initial conditions; dependent on you system

    T = 0.1
    nparticular = 0.08 / (RG.sigma**3 * np.pi / 6
                          )  # using finite differences df_dn
    # nparticular = 0.155/(RG.sigma**3*np.pi/6) # using analytic df_nd

    # T = 0.68
    # nparticular = 0.0414/(RG.sigma**3*np.pi/6)

    N = 20  # For publication plots, make this bigger (40? 80? 100? You decide!)
    Tc = 1.33
    Tlow = T

    # Bounds of minimization.
    # Use range that works for many temperatures
    # Left (vapor)
    a_vap = 1e-10 / (RG.sigma**3 * np.pi / 6)
    c_vap = nparticular
    # Right (liquid)
    a_liq = nparticular
    c_liq = 0.75 / (RG.sigma**3 * np.pi / 6)
    ###############################

    # Open file for output
    fout = open('figs/npart_RG-i%d-out.dat' % iterations, 'w')

    # label the columns of the output
    fout.write(
        '#T     nvapor     nliquid       phi(nvap)        phi(nliq)         nparticular\n'
    )  #       phi_avg')
    sys.stdout.flush()

    # Do first temperature before the loop
    nvapor, phi_vapor = minmax_RG.minimize(RG.phi, T, a_vap, c_vap,
                                           nparticular, iterations)
    print 'initial nvap,phi_vap', nvapor, phi_vapor
    nliquid, phi_liquid = minmax_RG.minimize(RG.phi, T, a_liq, c_liq,
                                             nparticular, iterations)
    print 'initial nliq,phi_liq', nliquid, phi_liquid
    print 'initial npart', nparticular * (RG.sigma**3 * np.pi / 6)

    #while T < 1.4:
    for j in xrange(0, N + 1):
        T = (Tc - Tlow) * (1 - ((N - j) / N)**4) + Tlow

        fout.flush()
        # Starting point for new nparticular is abscissa of max RG.phi with old nparticular
        nparticular = minmax_RG.maximize(RG.phi, T, nvapor, nliquid,
                                         nparticular, iterations)

        # I'm looking at the minima of RG.phi
        c_vap = nparticular
        a_liq = nparticular

        nvapor, phi_vapor = minmax_RG.minimize(RG.phi, T, c_vap, a_vap,
                                               nparticular, iterations)
        nliquid, phi_liquid = minmax_RG.minimize(RG.phi, T, a_liq, c_liq,
                                                 nparticular, iterations)

        tol = 1e-5
        fnpart = RG.phi(T, nparticular, nparticular, iterations)

        # Compare the two minima in RG.phi
        while np.fabs(phi_vapor - phi_liquid) / np.fabs(fnpart) > tol:

            delta_mu = (phi_liquid - phi_vapor) / (nliquid - nvapor)

            def newphi(T, n, npart, i):
                return RG.phi(T, n, npart, i) - delta_mu * n

            nparticular = minmax_RG.maximize(newphi, T, nvapor, nliquid,
                                             nparticular, iterations)

            fnpart = RG.phi(T, nparticular, nparticular, iterations)

            nvapor, phi_vapor = minmax_RG.minimize(RG.phi, T, a_vap, c_vap,
                                                   nparticular, iterations)
            nliquid, phi_liquid = minmax_RG.minimize(RG.phi, T, a_liq, c_liq,
                                                     nparticular, iterations)

        fout.write(str(T))
        fout.write('  ')
        fout.write(str(nvapor))
        fout.write('  ')
        fout.write(str(nliquid))
        fout.write('  ')
        fout.write(str(phi_vapor))
        fout.write('  ')
        fout.write(str(phi_liquid))
        fout.write('  ')
        fout.write(str(nparticular))
        fout.write('\n')
        sys.stdout.flush()
        print '   T, etaVap, etaLiq', T, nvapor / (
            RG.sigma**3 * np.pi / 6), nliquid / (RG.sigma**3 * np.pi / 6)
    fout.close()