def main(): print "Reading data...", print " %s."% options.rawFN, raw = rd( options.rawFN ) print " %s." % options.projFN, P = Serializer.LoadFromHDF( options.projFN ) print " %s." % options.msmFN msmDict = Serializer.LoadFromHDF( options.msmFN ) lens = [ ( P['TrajLengths'][i], i ) for i in range( P['TrajLengths'].shape[0] ) ] lens = sorted(lens,reverse=True) print lens trajList = [ b for (a,b) in lens[:options.nTraj] ] print trajList # Generate the trajectory data: trajData = [] for traj in trajList: trajData.append( array(rawObs( traj, P['TrajLengths'], raw ) ) ) # Generate the autocorrelations: print "Calculating the convolutions ... " trajCorr = [ autocorrelate.fft_autocorrelate( thing ) for thing in trajData ] print "Plotting the data ... " plotData( msmDict, trajCorr, trajList )
def main():
# First read the data:
fc = rd( options.fcFN )
dataList = []
for FN in options.dataFNs:
temp = rd(FN)
if len(temp.shape) !=1:
temp = temp[:,options.dataCol]
dataList.append( temp )
TSind = where( (0.5-options.eps<=fc)*(fc<=0.5+options.eps) )[0]
wd( ['TSE',str(options.eps) ], TSind )
avgList = []
stdList = []
# if dataList:
# for item in dataList:
# avgList.append( item[TSind].mean() )
# stdList.append( item[TSind].std() )
# avgList = array(avgList)
# stdList = array(stdList)
nameFN = wd( ['TSEaverages','Names'], array(dataList) )
# avgFN = wd( ['TSEaverages','Averages','eps-%.2f'%options.eps], avgList )
# stdFN = wd( ['TSEaverages','StdDevs','eps-%.2f'%options.eps], stdList )
figure()
width = 0.5
#bar( arange(len(avgList))+width/2., avgList,width, yerr=stdList, color='red',alpha=0.5,ecolor='black',edgecolor='black')
boxplot( [ item[TSind] for item in dataList ] )
barNames = [ '.'.join( thing.split('_')[-1].split('.')[:-2] ) for thing in options.dataFNs ]
title('TSE State Averages Pfold=0.5%s %.2f' % ( u"\u00B1", options.eps) )
ylabel('Fraction Native Contacts')
xticks( arange(1,len(options.dataFNs)+1), barNames, rotation=60 )
ylim([0,1])
savefig('TSEaverages_box_eps-%.2f.pdf'%options.eps)
parser.add_option('-x',dest='xFN',help='This is the x-coordinates for all the states')
parser.add_option('-y',dest='yFN',help='This is the y-coordinates for all the states')
parser.add_option('--xc',dest='xcol',type=int,help='The column in the x-data to use')
parser.add_option('--yc',dest='ycol',type=int,help='The column in the y-data to use')
parser.add_option('-d',dest='dataFN',help='Data to plot for all the states.')
parser.add_option('--pos-or-neg',dest='pos_or_neg',help='Pass this flag if you want to change the data to 1 and 0 corresponding to whether it is positive or negative. Like if you are plotting an eigenvector of a transition probability matrix.', default=False, action='store_true')
options,args = parser.parse_args()
import matplotlib
matplotlib.use('Pdf')
from matplotlib.pyplot import *
from numpy import *
from pyschwancr.dataIO import readData as rd
from pyschwancr.dataIO import writeData as wd
X = rd( options.xFN ).real
Y = rd( options.yFN ).real
data = rd( options.dataFN ).real
print X.shape, Y.shape, data.shape
if len( X.shape ) > 1:
print "X formatted strangely... Need to choose a column"
if not options.xcol:
print "Need to input the column to use as --xc!"
print "Continuing using column 0. Careful this could give wacky results!"
X = X[:,0]
else:
X = X[:,options.xcol]
if len( Y.shape ) > 1:
print "Y formatted strangely... Need to choose a column"
if not options.ycol:
def main():
print "Reading data...",
print " %s."% options.rawFN,
raw = rd( options.rawFN )
print " %s." % options.msmFN,
msm = rd( options.msmFN )
msm = msm[:,1] # Use the center column (This is how my data is formated.)
print " %s." % options.AsFN,
As = Serializer.LoadData( options.AsFN )
print " %s." % options.projFN,
P = Serializer.LoadFromHDF( options.projFN )
print " %s." % options.tFN
T = mmread( options.tFN )
lens = [ ( P['TrajLengths'][i], i ) for i in range( P['TrajLengths'].shape[0] ) ]
lens = sorted(lens,reverse=True)
trajList = [ b for (a,b) in lens[:options.nTraj] ]
rawFmt = ones( As.shape ) * -1
tempSum = 0
print raw.shape
for i in range( len( P['TrajLengths'] ) ):
rawFmt[i][ : P['TrajLengths'][i] ] = raw[ tempSum : tempSum + P['TrajLengths'][i] ]
tempSum += P['TrajLengths'][i]
# Generate the trajectory data:
trajData = []
msmData = []
N = int( P['TrajLengths'].max() / options.lag ) + 1
for traj in trajList:
print "\nCalculating trajectory %d" % traj,
print "Raw data.",
trajData.append( array(rawObs( traj, P['TrajLengths'], raw ) ) )
print "MSM Average."
start = zeros( len(msm) )
start[ As[ traj ][0] ] += 1
Cor, Traj, ObsTraj = Correlation.RawMSMCorrelation( T, rawFmt, As, Steps = N, StartingState = As[ traj ][0] )
msmData.append( Cor )
# Generate the autocorrelations:
print "Calculating the convolutions ... "
trajCorr = [ autocorrelate.fft_autocorrelate( thing ) for thing in trajData ]
msmCorr = [ autocorrelate.fft_autocorrelate( thing ) for thing in msmData ]
msmCorr = [ thing for thing in msmData ]
for i in range(1000):
if not os.path.exists('Autocorr%d'%i):
outDir = 'Autocorr%d'%i
os.mkdir( outDir )
break
if not outDir:
print "You have a lot of Autocorr directories..."
print "Writing output to ./ which could overwrite other data!"
outDir = './'
print "Writing the autocorrelation data"
for index, traj in enumerate( trajList ):
wd( [ 'autocorr','traj%d'%traj,'msm','lag%d'%options.lag ], msmCorr[index], dir=outDir )
wd( [ 'autocorr','traj%d'%traj,'raw'], trajCorr[index], dir=outDir )
plotData( msmCorr, trajCorr, trajList, options.lag )
