Tools for performing 1-component histogram reweighting for micellizaiton
Process
-----------------
(0) compile *f90 files (ifort filename.f90 filename.x)
(1) generate histograms using hisFile.py
it is a good idea to use a bash script like genHis.sh
(2) write input_hs.dat
this can also be done using a bash script like genHis.sh
(3) visualize the histograms using hisFile.py
(4) run order.x (/FILELOCATION/order.x)
(4) run entropy.x (/FILELOCATION/entropy.x)
(5) modify tolerance in entr2_par.dat
(6) run entropy2.x (/FILELOCATION/entropy2.x) (it generates prediction of partition function in pvt.dat file)
(7) compare the partition function predicition and the runs given with simulationVpartition.py
MICELLIZATION
(8) generate a a list of "T mu N" to create the lnZ-N curves with simulationVpartition.py
(9) find the CMC and see if the ideal-gas law (lnZ=N) is obeyed using partition2pressure.py
PHASE EQUILIBRIA
(8) modify entr4_par.dat values
(9) run entropy4.x (/FILELOCATION/entropy4.x)
(10) check the prediction of crit.dat vs. ising.dat
(11) modify entr2_par.dat 'smix' and 'peak separation' values based on entropy4 results
(12) run entropy2.x with the phase equilibria function (first: T_C mu_C -N_C then: T_min T_hi dT; where _C is the critical value)
Pre-analysis
------------------------------
hisFile.py
python program that handles his*.dat files
get help:
hisFile.py -h
to generate histogram from cassandra/non-AZP simulations
usage:
hisFile.py -r '$NAME' -w $width -o cassandra -s $skip
where:
$NAME: is from $NAME.box1.prp1
$width: Energy bin width
$skip: # MC steps to skip
to plot the histograms
usage:
hisFile.py -i input_hs.dat -p -s
"-p": to plot histograms
"-s": show the histograms using matplotlib
Histogram reweighting programs
------------------------------
entropy.f90 [must use first]
histogram reweighting via a single pass w/o Ferrenbery-Swendsen algorithm
generates input_hs2.dat
input_hs.dat
list of histograms
Name structure: his$NAME$ending.dat
line 1: #_$ending_types
line 2 - #_$ending_types: $ending
(The program was set up for simulations that created repeat
sims that had different character ending in the filename.)
line 3 - EOF : $NAME
order.f90
Orders the files in input_hs.dat so that there is a maximum initial overlap
entropy2.f90
Ferrenbery-Swendsen algorithm program
input_hs2.dat
Input for entropy2, made by entropy, modified by entropy2 and entropy4
entr2_par.dat
sets tolerance for Ferrenbery-Swendsen algorithm
Only need to worry about the tolerance for micellization
line 1:
location of ising.dat
line 2:
'muincr' incremental mu value for F-S algorithm optimization (sensitive for the time)
'2D' T for 2d sims, F for 3d sims
'read_s' T for read in the smix value for critical point from command line, F for read from file
line 3:
'tolerance' tolerance for converged histograms
'accfactor' acceleration factor for F-S algorithm optimization
line 4:
'mindens' minimum accepted peak separation in ising order parameter
line 5:
'smix' s parameter in universality scaling
'ncrit' number of points in the critical point order parameter, the higher the more points in crit.dat
ising.dat
Ising model data for phase separation analysis.
put in the same folder as input_hs2.dat or change entr2_par.dat
entropy4.f90
Not for Micellization, but for thermodynamic phase transitions
entr4_par.dat
parameters for entropy4
line 1:
sets the initial vector for searching for the optimum where rho_l = rho_v (integral under the gas and liquid density distribution are equal)
'T1' initial direction and mag for changing the guess for T_C if the first histogram in input_hs.dat is below the T_C make T1>0, other wise, T1<0
'mu1' initial direction and mag for changing the guess for mu_C if the first histogram in input_hs.dat is below the mu_C make mu1>0, other wise, mu1<0
'smix1' initial direction and mag for changing the guess for smix_C if the first histogram in input_hs.dat is below the smix_C make smix1>0, other wise, smix1<0
line 2:
'ftol' tolerance for convergence of optimization
'dmu_dT' Change of mu vs. T at rho = rho_c, estimate from the mu and T range of your histograms
'smix' s field mixing parameter initial guess
Micellization Post-analysis
------------------------------
simulationVpartition.py
python program to compare the <N> and <E> from simulation histograms with the
partition function calculated from those histograms
usage:
python simulationVpartition.py -i input_hs.dat
partition2pressure.py
python program to calculate the Pressure versus concentration
usage:
python partition2pressure.py -c -l $L -p -o -s -m curvature
$L : box length [Angstrom]
"m" : define the method
"c" : concentration on the x-axis [mM], pressure [kPa]
"o" : output the Pressure versus concentration
"p" : plot the curve (pvt.png)
"s" : show the curve