| 1 | #!/usr/bin/env python | 
| 2 | """Pressure Correlation function | 
| 3 |  | 
| 4 | Computes various correlation functions of the pressure and pressure tensor | 
| 5 | that have been stored in a stat file.   These can be used to compute | 
| 6 | shear and bulk viscosities. | 
| 7 |  | 
| 8 | Usage: stat2visco | 
| 9 |  | 
| 10 | Options: | 
| 11 | -h, --help              show this help | 
| 12 | -f, --stat-file=...     use specified stat file | 
| 13 | -o, --output-file=...   use specified output (.pcorr) file | 
| 14 | -g, --green-kubo        use Green-Kubo formulae (noisy!) | 
| 15 | -e, --einstein          use Einstein relation (best) | 
| 16 | -s, --shear             compute the shear viscosity (the off-diagonal | 
| 17 | pressure tensor values must be present in the .stat | 
| 18 | file) | 
| 19 |  | 
| 20 | The Green-Kubo formulae option will compute: V*<(P(t)-<P>)*(P(0)-<P>)>/kT , | 
| 21 | which may be integrated to give a slowly-converging value for the viscosity. | 
| 22 |  | 
| 23 | The Einstein relation option will compute: V*<(\int_0^t (P(t')-<P>)dt')^2>/2kT, | 
| 24 | which will grow approximately linearly in time.  The long-time slope of this | 
| 25 | function will be the viscosity. | 
| 26 |  | 
| 27 | Example: | 
| 28 | stat2visco -f ring5.stat -o ring5.pcorr -e -s | 
| 29 |  | 
| 30 | """ | 
| 31 |  | 
| 32 | __author__ = "Dan Gezelter (gezelter@nd.edu)" | 
| 33 | __version__ = "$Revision: 1.1 $" | 
| 34 | __date__ = "$Date: 2008-01-23 21:22:18 $" | 
| 35 |  | 
| 36 | __copyright__ = "Copyright (c) 2007 by the University of Notre Dame" | 
| 37 | __license__ = "OOPSE" | 
| 38 |  | 
| 39 | import sys | 
| 40 | import getopt | 
| 41 | import string | 
| 42 | import math | 
| 43 |  | 
| 44 | def usage(): | 
| 45 | print __doc__ | 
| 46 |  | 
| 47 | def readStatFile(statFileName): | 
| 48 |  | 
| 49 | global time | 
| 50 | global temperature | 
| 51 | global pressure | 
| 52 | global volume | 
| 53 | time = [] | 
| 54 | temperature = [] | 
| 55 | pressure = [] | 
| 56 | volume = [] | 
| 57 |  | 
| 58 | if (doShear): | 
| 59 | global Pxx | 
| 60 | global Pyy | 
| 61 | global Pzz | 
| 62 | global Pxy | 
| 63 | global Pxz | 
| 64 | global Pyz | 
| 65 |  | 
| 66 | Pxx = [] | 
| 67 | Pyy = [] | 
| 68 | Pzz = [] | 
| 69 | Pxy = [] | 
| 70 | Pxz = [] | 
| 71 | Pyz = [] | 
| 72 |  | 
| 73 | statFile = open(statFileName, 'r') | 
| 74 | line = statFile.readline() | 
| 75 |  | 
| 76 | print "reading File" | 
| 77 | pressSum = 0.0 | 
| 78 | volSum = 0.0 | 
| 79 | tempSum = 0.0 | 
| 80 | line = statFile.readline() | 
| 81 | while 1: | 
| 82 | L = line.split() | 
| 83 | time.append(float(L[0])) | 
| 84 | temperature.append(float(L[4])) | 
| 85 | # | 
| 86 | # OOPSE prints out pressure in units of atm. | 
| 87 | # | 
| 88 | pressure.append(float(L[5])) | 
| 89 | volume.append(float(L[6])) | 
| 90 |  | 
| 91 | if doShear: | 
| 92 | if (len(L) > 16): | 
| 93 | # | 
| 94 | # OOPSE prints out the pressure tensor in units of amu*fs^-2*Ang^-1 | 
| 95 | # | 
| 96 | Pxx.append(float(L[8])) | 
| 97 | Pyy.append(float(L[12])) | 
| 98 | Pzz.append(float(L[16])) | 
| 99 | # | 
| 100 | # symmetrize the off-diagonal terms in the pressure tensor | 
| 101 | # | 
| 102 | Pxy.append(0.5*(float(L[9])  + float(L[11]))) | 
| 103 | Pxz.append(0.5*(float(L[10]) + float(L[14]))) | 
| 104 | Pyz.append(0.5*(float(L[13]) + float(L[15]))) | 
| 105 | else: | 
| 106 | print "Not enough columns are present in the .stat file" | 
| 107 | print "to calculate the shear viscosity..." | 
| 108 | print | 
| 109 | print "stat2visco expects to find all 9 elements of the" | 
| 110 | print "pressure tensor in columns 9-17 of the .stat file" | 
| 111 | print | 
| 112 | print "You may need to set the statFileFormat string" | 
| 113 | print "explicitly in your .md file when running OOPSE." | 
| 114 | print "Consult the OOPSE documentation for more details." | 
| 115 | sys.exit() | 
| 116 |  | 
| 117 | line = statFile.readline() | 
| 118 | if not line: break | 
| 119 |  | 
| 120 | statFile.close() | 
| 121 |  | 
| 122 | def computeAverages(): | 
| 123 |  | 
| 124 | global tempAve | 
| 125 | global pressAve | 
| 126 | global volAve | 
| 127 | global pvAve | 
| 128 |  | 
| 129 | print "computing Averages" | 
| 130 |  | 
| 131 | tempSum = 0.0 | 
| 132 | pressSum = 0.0 | 
| 133 | volSum = 0.0 | 
| 134 | pvSum = 0.0 | 
| 135 |  | 
| 136 | temp2Sum = 0.0 | 
| 137 | press2Sum = 0.0 | 
| 138 | vol2Sum = 0.0 | 
| 139 | pv2Sum = 0.0 | 
| 140 |  | 
| 141 | # converts amu*fs^-2*Ang^-1 -> atm | 
| 142 | pressureConvert = 1.63882576e8 | 
| 143 |  | 
| 144 | for i in range(len(time)): | 
| 145 | tempSum = tempSum + temperature[i] | 
| 146 | pressSum = pressSum + pressure[i] | 
| 147 | volSum = volSum + volume[i] | 
| 148 | # in units of amu Ang^2 fs^-1 | 
| 149 | pvTerm = pressure[i]*volume[i] / pressureConvert | 
| 150 | pvSum = pvSum + pvTerm | 
| 151 | temp2Sum = temp2Sum + math.pow(temperature[i],2) | 
| 152 | press2Sum = press2Sum + math.pow(pressure[i],2) | 
| 153 | vol2Sum = vol2Sum + math.pow(volume[i],2) | 
| 154 | pv2Sum = pv2Sum + math.pow(pvTerm,2) | 
| 155 |  | 
| 156 | tempAve = tempSum / float(len(time)) | 
| 157 | pressAve = pressSum / float(len(time)) | 
| 158 | volAve = volSum / float(len(time)) | 
| 159 | pvAve = pvSum / float(len(time)) | 
| 160 |  | 
| 161 | tempSdev = math.sqrt(temp2Sum / float(len(time)) - math.pow(tempAve,2)) | 
| 162 | pressSdev = math.sqrt(press2Sum / float(len(time)) - math.pow(pressAve,2)) | 
| 163 | if (vol2Sum / float(len(time)) < math.pow(volAve,2)): | 
| 164 | volSdev = 0.0 | 
| 165 | else: | 
| 166 | volSdev = math.sqrt(vol2Sum / float(len(time)) - math.pow(volAve,2)) | 
| 167 | pvSdev = math.sqrt(pv2Sum / float(len(time)) - math.pow(pvAve,2)) | 
| 168 |  | 
| 169 | print "   Average pressure = %f +/- %f (atm)" % (pressAve, pressSdev) | 
| 170 | print "     Average volume = %f +/- %f (Angst^3)" % (volAve, volSdev) | 
| 171 | print "Average temperature = %f +/- %f (K)" % (tempAve, tempSdev) | 
| 172 | print " Average PV product = %f +/- %f (amu Angst^2 fs^-1)" % (pvAve, pvSdev) | 
| 173 |  | 
| 174 | def computeCorrelations(outputFileName): | 
| 175 |  | 
| 176 | # converts amu*fs^-2*Ang^-1 -> atm | 
| 177 | pressureConvert = 1.63882576e8 | 
| 178 |  | 
| 179 | # Boltzmann's constant amu*Ang^2*fs^-2/K | 
| 180 | kB = 8.31451e-7 | 
| 181 |  | 
| 182 | # converts amu Ang^-1 fs^-1  ->  g cm^-1 s^-1 | 
| 183 | viscoConvert = 0.16605387 | 
| 184 |  | 
| 185 | preV = viscoConvert * volAve / (kB * tempAve) | 
| 186 | preVi = viscoConvert / (volAve * kB * tempAve) | 
| 187 |  | 
| 188 | if doGreenKubo: | 
| 189 | gkPcorr = [] | 
| 190 | if doShear: | 
| 191 | gkXYcorr = [] | 
| 192 | gkXZcorr = [] | 
| 193 | gkYZcorr = [] | 
| 194 | print "computing Green-Kubo-style Correlation Function" | 
| 195 | # i corresponds to dt | 
| 196 | for i in range(len(time)): | 
| 197 | # j is the starting time for the correlation | 
| 198 | pp = 0.0 | 
| 199 | if doShear: | 
| 200 | ppXY = 0.0 | 
| 201 | ppXZ = 0.0 | 
| 202 | ppYZ = 0.0 | 
| 203 | for j in range( len(time) - i ): | 
| 204 | pv1 = pressure[j]*volume[j]/pressureConvert - pvAve | 
| 205 | pv2 = pressure[j+i]*volume[j+i]/pressureConvert - pvAve | 
| 206 | pp = pp + pv1*pv2 | 
| 207 | if doShear: | 
| 208 | ppXY = ppXY + Pxy[j+i]*Pxy[j] | 
| 209 | ppXZ = ppXZ + Pxz[j+i]*Pxz[j] | 
| 210 | ppYZ = ppYZ + Pyz[j+i]*Pyz[j] | 
| 211 |  | 
| 212 | gkPcorr.append(pp / float(len(time) - i)) | 
| 213 | if doShear: | 
| 214 | gkXYcorr.append(ppXY / float(len(time)-i)) | 
| 215 | gkXZcorr.append(ppXZ / float(len(time)-i)) | 
| 216 | gkYZcorr.append(ppYZ / float(len(time)-i)) | 
| 217 |  | 
| 218 |  | 
| 219 | if doEinstein: | 
| 220 | print "computing Einstein-style Correlation Function" | 
| 221 |  | 
| 222 | # Precompute sum variables to aid integration. | 
| 223 | # The integral from t0 -> t0 + t  can be easily obtained | 
| 224 | # from the precomputed sum variables:  sum[t0+t] - sum[t0-1] | 
| 225 | pSum = [] | 
| 226 | pSum.append( (pressure[0] - pressAve) / pressureConvert) | 
| 227 | for i in range(1, len(time)): | 
| 228 | pSum.append(pSum[i-1] + (pressure[i]-pressAve)/pressureConvert ) | 
| 229 |  | 
| 230 | if doShear: | 
| 231 | xySum = [] | 
| 232 | xySum.append(Pxy[0]) | 
| 233 | xzSum = [] | 
| 234 | xzSum.append(Pxz[0]) | 
| 235 | yzSum = [] | 
| 236 | yzSum.append(Pyz[0]) | 
| 237 | for i in range(1, len(time)): | 
| 238 | xySum.append(xySum[i-1] + Pxy[i]) | 
| 239 | xzSum.append(xzSum[i-1] + Pxz[i]) | 
| 240 | yzSum.append(yzSum[i-1] + Pyz[i]) | 
| 241 |  | 
| 242 |  | 
| 243 | ePcorr = [] | 
| 244 | dt = time[1] - time[0] | 
| 245 |  | 
| 246 | if doShear: | 
| 247 | eXYcorr = [] | 
| 248 | eXZcorr = [] | 
| 249 | eYZcorr = [] | 
| 250 |  | 
| 251 | # i corresponds to the total duration of the integral | 
| 252 | for i in range(len(time)): | 
| 253 | pIntSum = 0.0 | 
| 254 | if doShear: | 
| 255 | xyIntSum = 0.0 | 
| 256 | xzIntSum = 0.0 | 
| 257 | yzIntSum = 0.0 | 
| 258 | # j corresponds to the starting point of the integral | 
| 259 | for j in range(len(time) - i): | 
| 260 | if (j == 0): | 
| 261 | pInt = dt*pSum[j+i] | 
| 262 | if doShear: | 
| 263 | xyInt = dt*xySum[j+i] | 
| 264 | xzInt = dt*xzSum[j+i] | 
| 265 | yzInt = dt*yzSum[j+i] | 
| 266 | else: | 
| 267 | pInt = dt*(pSum[j+i] - pSum[j-1]) | 
| 268 | if doShear: | 
| 269 | xyInt = dt*(xySum[j+i] - xySum[j-1]) | 
| 270 | xzInt = dt*(xzSum[j+i] - xzSum[j-1]) | 
| 271 | yzInt = dt*(yzSum[j+i] - yzSum[j-1]) | 
| 272 |  | 
| 273 | pIntSum = pIntSum + pInt*pInt | 
| 274 | if doShear: | 
| 275 | xyIntSum = xyIntSum + xyInt*xyInt | 
| 276 | xzIntSum = xzIntSum + xzInt*xzInt | 
| 277 | yzIntSum = yzIntSum + yzInt*yzInt | 
| 278 | ePcorr.append(pIntSum / float(len(time)-i)) | 
| 279 | if doShear: | 
| 280 | eXYcorr.append(xyIntSum / float(len(time)-i)) | 
| 281 | eXZcorr.append(xzIntSum / float(len(time)-i)) | 
| 282 | eYZcorr.append(yzIntSum / float(len(time)-i)) | 
| 283 |  | 
| 284 |  | 
| 285 | outputFile = open(outputFileName, 'w') | 
| 286 | for i in range(len(time)): | 
| 287 | if doGreenKubo: | 
| 288 | if doShear: | 
| 289 | outputFile.write("%f\t%13e\t%13e\t%13e\t%13e\n" % (time[i], preVi*gkPcorr[i], preV*gkXYcorr[i], preV*gkXZcorr[i], preV*gkYZcorr[i])) | 
| 290 | else: | 
| 291 | outputFile.write("%f\t%13e\n" % (time[i], preVi*gkPcorr[i])) | 
| 292 |  | 
| 293 | if doEinstein: | 
| 294 | if doShear: | 
| 295 | outputFile.write("%f\t%13e\t%13e\t%13e\t%13e\n" % (time[i], 0.5*preV*ePcorr[i], 0.5*preV*eXYcorr[i], 0.5*preV*eXZcorr[i], 0.5*preV*eYZcorr[i])) | 
| 296 | else: | 
| 297 | outputFile.write("%f\t%13e\n" % (time[i], 0.5*preV*ePcorr[i])) | 
| 298 | outputFile.close() | 
| 299 |  | 
| 300 | def main(argv): | 
| 301 | global doGreenKubo | 
| 302 | global doEinstein | 
| 303 | global doShear | 
| 304 | global haveStatFileName | 
| 305 | global haveOutputFileName | 
| 306 |  | 
| 307 | haveStatFileName = False | 
| 308 | haveOutputFileName = False | 
| 309 | doShear = False | 
| 310 | doGreenKubo = False | 
| 311 | doEinstein = False | 
| 312 |  | 
| 313 | try: | 
| 314 | opts, args = getopt.getopt(argv, "hgesf:o:", ["help", "green-kubo", "einstein", "shear", "stat-file=", "output-file="]) | 
| 315 | except getopt.GetoptError: | 
| 316 | usage() | 
| 317 | sys.exit(2) | 
| 318 | for opt, arg in opts: | 
| 319 | if opt in ("-h", "--help"): | 
| 320 | usage() | 
| 321 | sys.exit() | 
| 322 | elif opt in ("-g", "--green-kubo"): | 
| 323 | doGreenKubo = True | 
| 324 | elif opt in ("-e", "--einstein"): | 
| 325 | doEinstein = True | 
| 326 | elif opt in ("-s", "--shear"): | 
| 327 | doShear = True | 
| 328 | elif opt in ("-f", "--stat-file"): | 
| 329 | statFileName = arg | 
| 330 | haveStatFileName = True | 
| 331 | elif opt in ("-o", "--output-file"): | 
| 332 | outputFileName = arg | 
| 333 | haveOutputFileName = True | 
| 334 | if (not haveStatFileName): | 
| 335 | usage() | 
| 336 | print "No stat file was specified" | 
| 337 | sys.exit() | 
| 338 | if (not haveOutputFileName): | 
| 339 | usage() | 
| 340 | print "No output file was specified" | 
| 341 | sys.exit() | 
| 342 |  | 
| 343 | readStatFile(statFileName); | 
| 344 | computeAverages(); | 
| 345 | computeCorrelations(outputFileName); | 
| 346 |  | 
| 347 | if __name__ == "__main__": | 
| 348 | if len(sys.argv) == 1: | 
| 349 | usage() | 
| 350 | sys.exit() | 
| 351 | main(sys.argv[1:]) |