def test_2(self):
try:
subprocess.Popen('ipcluster start --cluster-id=testclusterid --n=1 --daemonize', shell=True)
time.sleep(5)
args = self.Args()
args.output_dir = tempfile.mkdtemp()
args.cluster_id = 'testclusterid'
logger = AssignParallel.setup_logger()
AssignParallel.main(args, self.metric, logger)
assignments = Serializer.load_data(os.path.join(args.output_dir, 'Assignments.h5'))
r_assignments = Serializer.load_data(os.path.join(fixtures_dir(), 'Assignments.h5'))
distances = Serializer.load_data(os.path.join(args.output_dir, 'Assignments.h5.distances'))
r_distances = Serializer.load_data(os.path.join(fixtures_dir(), 'Assignments.h5.distances'))
npt.assert_array_equal(assignments, r_assignments)
npt.assert_array_almost_equal(distances, r_distances)
except:
raise
finally:
shutil.rmtree(args.output_dir)
subprocess.Popen('ipcluster stop', shell=True).wait()
def test_2(self):
try:
subprocess.Popen(
'ipcluster start --cluster-id=testclusterid --n=1 --daemonize',
shell=True)
time.sleep(5)
args = self.Args()
args.output_dir = tempfile.mkdtemp()
args.cluster_id = 'testclusterid'
logger = AssignParallel.setup_logger()
AssignParallel.main(args, self.metric, logger)
assignments = Serializer.load_data(
os.path.join(args.output_dir, 'Assignments.h5'))
r_assignments = Serializer.load_data(
os.path.join(fixtures_dir(), 'Assignments.h5'))
distances = Serializer.load_data(
os.path.join(args.output_dir, 'Assignments.h5.distances'))
r_distances = Serializer.load_data(
os.path.join(fixtures_dir(), 'Assignments.h5.distances'))
npt.assert_array_equal(assignments, r_assignments)
npt.assert_array_almost_equal(distances, r_distances)
except:
raise
finally:
shutil.rmtree(args.output_dir)
subprocess.Popen('ipcluster stop', shell=True).wait()
def save_container(filename, dtype):
s = Serializer({
'Data': np.array(minus_ones, dtype=dtype),
'completed_vtrajs': np.zeros((n_vtrajs), dtype=np.bool),
'hashes': hashes
})
s.save_to_hdf(filename)
def __init__(self, information, projectfile, populationfile,
assignmentfile_fixed, tmatrixfile, rawdatafile):
try:
self.Info = information
self.ProjectInfo = Serializer.LoadFromHDF(projectfile)
self.Population = loadtxt(populationfile)
self.Assignments = Serializer.LoadFromHDF(assignmentfile_fixed)
self.Tmatrix = mmread(tmatrixfile)
self.StateAssignment = hct.get_StatesAssignments(self.Assignments)
self.getrawdata(rawdatafile)
except:
print "Having trouble with getting required files"
raise
def CalculateProjectRg(ProjectInfo, Output, returnRgs=False):
"""
Calculate Radius of gyration for the Project ie. all the Trajectories.
ProjectInfo: ProjectInfo.h5 file.
Output: output file (XXX.dat).
The Output default will be set in the scripts and it is './Rgs.dat'.
"""
Output = checkoutput(Output)
if not isinstance(ProjectInfo, str):
print "Please input the Path to ProjectInfo.h5"
raise IOError
print 'Calculating the Rg for each trajectory......'
ProjectInfoPath = '/'.join(os.path.realpath(ProjectInfo).split('/')[:-1])
os.chdir(ProjectInfoPath)
Trajfiles = []
ProjectInfo = Serializer.LoadFromHDF(ProjectInfo)
for i in range(ProjectInfo['NumTrajs']):
Trajfiles.append(ProjectInfo['TrajFilePath'] +
ProjectInfo['TrajFileBaseName'] + '%d' % i +
ProjectInfo['TrajFileType'])
Rgs = computeRg(Trajfiles)
print "Save data to %s" % Output
savetxt(Output, Rgs)
print "Done."
if returnRgs:
return Rgs
def main():
parser = argparse.ArgumentParser()
parser.add_argument('-p', '--project', default='ProjectInfo.h5')
parser.add_argument('-a', '--assignments', default='Data/Assignments.h5')
args = parser.parse_args()
a = Serializer.LoadData(args.assignments)
p = Project.LoadFromHDF(args.project)
maxx, maxy, minx, miny = -np.inf, -np.inf, np.inf, np.inf
n_states = np.max(a) + 1
x = np.concatenate([p.LoadTraj(i)['XYZList'][:, 0, 0] for i in range(p['NumTrajs'])])
y = np.concatenate([p.LoadTraj(i)['XYZList'][:, 0, 1] for i in range(p['NumTrajs'])])
a = np.concatenate([a[i, :] for i in range(p['NumTrajs'])])
plot_v(minx=np.min(x), maxx=np.max(x), miny=np.min(y), maxy=np.max(y))
colors = ['b', 'r', 'm', 'c', 'g']
for j in xrange(n_states):
w = np.where(a == j)[0]
pp.scatter(x[w], y[w], marker='x', c=colors[j], label='State %d' % j,
edgecolor=colors[j], alpha=0.5)
pp.legend()
pp.show()
def save_new_ref(filename, data):
""" Saves a new version of the reference data, and backs up the old """
ext = filename.split('.')[-1]
if (data == None):
print "WARNING: Error generating file: %s" % filename
print "Skipped... try again."
return
if os.path.exists(filename):
os.system('mv %s %s' % (filename, BACKUP_DIR))
if ext in ['h5', 'lh5']:
if scipy.sparse.issparse(data):
data = data.toarray()
Serializer.SaveData(filename, data)
elif ext == 'mtx':
io.mmwrite(filename, data)
elif ext == 'pkl':
f = open(filename, 'w')
pickle.dump(f, data)
f.close()
else:
raise ValueError('Could not understand extension (.%s) for %s' %
(ext, filename))
return
def check_container(filename):
ondisk = Serializer.load_from_hdf(filename)
if n_vtrajs != len(ondisk['hashes']):
raise ValueError('You asked for {} vtrajs, but your checkpoint \
file has {}'.format(n_vtrajs, len(ondisk['hashes'])))
if not np.all(ondisk['hashes'] == hashes):
raise ValueError('Hash mismatch. Are these checkpoint files for \
the right project?')
def get_projectinfo():
try:
projectinfo = Serializer.LoadFromHDF('../ProjectInfo.h5')
except IOError:
print "Can't find ProjectInfo.h5!"
return projectinfo
def check_container(filename):
ondisk = Serializer.load_from_hdf(filename)
if n_vtrajs != len(ondisk['hashes']):
raise ValueError('You asked for {} vtrajs, but your checkpoint \
file has {}'.format(n_vtrajs, len(ondisk['hashes'])))
if not np.all(ondisk['hashes'] ==
hashes):
raise ValueError('Hash mismatch. Are these checkpoint files for \
the right project?')
def get_Trajectory_frame(trajid,frames):
"""
Get trajectory frames.
From Trajectory file(traj_.lh5) get the frame.
"""
Path = "/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/Trajectories"
traj = Serializer.LoadFromHDF('%s/trj%d.lh5'%(Path,trajid))
return [traj['XYZList'][i] for i in frames]
def pseudosampling(states, assignment, numberoftrajs, frames, output):
try:
fn = Serializer.LoadFromHDF(assignment)
except IOError:
print "Can't find Assignment file"
sys.exit()
for stateid in states:
a = pseudotrajs(numberoftrajs, stateid, fn, frames)
a.SaveToHDF(output)
print "Wrote:%s" % output
def calculate_statepopulation_rawdata(AssignmentsFixed):
a = Serializer.LoadFromHDF(AssignmentsFixed)
statenumber = max([max(a['Data'][i]) for i in range(len(a['Data']))]) + 1
p = np.zeros(statenumber)
for state in range(statenumber):
for traj in range(len(a['Data'])):
p[state] += a['Data'][traj].tolist().count(state)
p = p / p.sum()
return p
def getrawdata(self, rawdatafile):
try:
d = Serializer.LoadFromHDF(rawdatafile)
self.RawData = d['Data']
except:
try:
self.RawData = loadtxt(rawdatafile)
except:
print "Can not load {}".format(rawdatafile)
raise
def get_StatesAssignments(AssignmentFiles):
"""
StatesAssignments {'state':{'Trajectory':[Frames]}}
"""
A = AssignmentFiles
if isinstance(A,str):
A = Serializer.LoadFromHDF(A)
S = {}
try:
for trajid,data in zip(A['TrajID'],A['Data']):
for i in range(len(data)):
if data[i] == -1:
continue
else:
S.setdefault('%d'%data[i],{}).setdefault('%d'%trajid,[]).append(i)
return S
except KeyError:
# Assignments file which doesn't has key 'TrajID' should be regular Assignments file instead of new assignments files created for bootstrap. Then modify the assignments file i.e, create key 'TrajID'.
A['TrajID'] = list(range(len(A['Data'])))
return get_StatesAssignments(A)
def test():
assignments = Serializer.LoadFromHDF("/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/RMSDCluster4.2/Data/Assignments.h5")
StatesAsi = get_StatesAssignments(assignments)
NumHelix_states = compute_numhelix_states(StatesAsi)
#print "NumHelix_states",NumHelix_states
#savetxt('NumHelix_states',NumHelix_states)
states = [int(i) for i in NumHelix_states.keys()]
states.sort()
mean_numhelix_states = []
std_numhelix_states = []
for state in states:
mean_numhelix_states.append(np.mean(NumHelix_states['%d'%state]))
std_numhelix_states.append(np.std(NumHelix_states['%d'%state]))
plt.figure()
plt.errorbar(states,mean_numhelix_states,std_numhelix_states)
plt.xlabel("State ID")
plt.ylabel("Number of Helix")
plt.savefig("Numhelix_states")
plt.show()
def bootstrap(AssignmentsFile,numtraj,bootstrapnumber,PathtoSaveFiles):
bootstraplist,TrajID = [],[]
File = Serializer.LoadFromHDF(AssignmentsFile)
datalist = File['Data']
if isinstance(numtraj,str) and numtraj.lower()=='all':
numtraj = len(datalist)
elif int(numtraj) <= 0:
print "Please input valid number from 1 to %d"%len(datalist)
sys.exit()
else:
numtraj = int(numtraj)
for i in range(bootstrapnumber):
all_assignments = -1 * np.ones((numtraj,len(datalist[0])), dtype=np.int)
k = 0
trajid = []
for j,m in sample_with_replacement(datalist,numtraj):
all_assignments[k][:] = j[:]
trajid.append(m)
k += 1
TrajID.append(trajid)
bootstraplist.append(all_assignments)
SaveBootstrapFiles(bootstraplist,TrajID,PathtoSaveFiles,File,AssignmentsFile)
def compute_numhelix_states(StatesAssignments):
"""
Compute the average number of helix for all states.
Need Path to trj_hc.h5
"""
SA = StatesAssignments
states = SA.keys()
numhelix_states = {}
n = 0
for state in states:
n +=1
print "Compute number of helix for state %d/%d"%(n,len(states))
TrajID = SA[state].keys()
numhelix_state = []
for trajid in TrajID:
T = {}
TrajFile = '/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/Trajectories/trj%s_hc.h5'%trajid
Traj = Serializer.LoadFromHDF(TrajFile)
T['HCs'] = [Traj['HCs'][i] for i in SA[state][trajid]]
numhelix_state += count_Helix(T)
numhelix_states[state] = numhelix_state
return numhelix_states
x = (t['XYZList'][i, Atom1, :] - t['XYZList'][i, Atom2, :])[0]
x = x.tolist()
distance.append(np.dot(x, x)**0.5)
distance += [-1] * (LongestTrajLength - len(t['XYZList']))
return distance
#------------MAIN---------------
AtomName1 = 'C'
ResidueID1 = 1
AtomName2 = 'N'
ResidueID2 = 23
path = '/Users/tud51931/projects/MSM/msm/ff03ERR-hybridkcenter/RMSDCluster4.0'
Distances = []
ProjectInfo = Serializer.LoadFromHDF('%s/ProjectInfo.h5' % path)
LongestTrajLength = max(ProjectInfo['TrajLengths'])
os.chdir(path)
if os.path.exists('EndtoEndDistances.dat'):
print "EndtoEndDistances.dat exists!"
sys.exit()
print 'Calculating the eeDistance of each trajectory......'
for i in range(ProjectInfo['NumTrajs']):
trajfile = ProjectInfo['TrajFilePath'] + ProjectInfo[
'TrajFileBaseName'] + '%d' % i + ProjectInfo['TrajFileType']
print '%d in %d Trajectories' % (i, ProjectInfo['NumTrajs']), trajfile
d = calculatedistance(AtomName1, ResidueID1, AtomName2, ResidueID2,
trajfile, LongestTrajLength)
Distances.append(d)
print "Save data to ./EndtoEndDistance.dat"
savetxt('EndtoEndDistances.dat', Distances)
from msmbuilder import Serializer
def draw_index(probs, n_picks=1, UseFastMethod=True):
"""Draw a number (or many numbers, controlled by n_picks), weighted by the probabilities probs."""
if UseFastMethod:
t = np.cumsum(probs)
s = sum(probs)
return np.searchsorted(t, np.random.rand(n_picks) * s)
tcounts = mmread(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/RMSDCluster4.2/lagtime50/tCounts.UnMapped.mtx'
)
Assignment = Serializer.LoadFromHDF(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/RMSDCluster4.2/lagtime50/Assignments.Fixed.h5'
)
trajnum = 100
frames = Assignment['Data'].shape[1]
a = Serializer()
a['Data'] = -1 * np.ones((trajnum, frames))
for traj in range(trajnum):
print '%d of %d Trajectories' % (traj, trajnum)
startstate = 126
a['Data'][traj, 0] = startstate
for step in range(1, frames):
probs = tcounts.data[tcounts.row == startstate] / sum(
tcounts.data[tcounts.row == startstate])
a['Data'][traj, step] = tcounts.col[tcounts.row == startstate][
draw_index(probs)[0]]
startstate = a['Data'][traj, step]
the structures of each of the cluster centers. Produced using Cluster.py.''',
default='Data/Gens.lh5')
parser.add_argument('output_dir', default='PDBs')
args = parser.parse_args()
if -1 in args.states:
print "Ripping PDBs for all states"
args.states = 'all'
if args.conformations_per_state == -1:
print "Getting all PDBs for each state"
args.conformations_per_state = 'all'
atom_indices = np.loadtxt(args.lprmsd_atom_indices, np.int)
assignments = Serializer.LoadData(args.assignments)
project = Project.load_from_hdf(args.project)
if args.lprmsd_permute_atoms == 'None':
permute_indices = None
else:
permute_indices = ReadPermFile(args.lprmsd_permute_atoms)
if args.lprmsd_alt_indices == 'None':
alt_indices = None
else:
alt_indices = np.loadtxt(args.lprmsd_alt_indices, np.int)
run(project, assignments, args.conformations_per_state, args.states,
args.output_dir, args.generators, atom_indices, permute_indices,
alt_indices, args.total_memory_gb)
help="Input RMSD.h5 file",
required=True)
parser.add_argument('-rg', '--Rg', help="Input Rgs.dat file", required=True)
parser.add_argument('-l',
'--Locate',
help="Locate states on the Rg-RMSD graph",
type=str)
parser.add_argument('-o',
'--Output',
help="Output file (graph) name.Default: Rg-RMSD.png",
default="Rg-RMSD.png")
args = parser.parse_args()
try:
R = Serializer.LoadFromHDF(args.RMSD)
rmsd = []
for i in range(len(R['Data'])):
for j in range(len(R['Data'][i])):
if R['Data'][i, j] != -1:
rmsd.append(R['Data'][i, j])
except IOError:
print "Can't find RMSD.h5, please run CalculateProjectRMSD.py first to get RMSD.h5."
raise IOError
try:
Rgs = loadtxt(args.Rg)
rgs = []
for i in range(len(Rgs)):
for j in range(len(Rgs[i])):
if Rgs[i, j] != -1:
rgs.append(Rgs[i, j])
import os, sys
import numpy as np
from msmbuilder import Serializer, Trajectory
import matplotlib.pyplot as plt
sys.path.append("/Users/tud51931/scripts/gfzhou")
import HelixCoilTools as hct
ProjectInfo = Serializer.LoadFromHDF(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/ProjectInfo.h5'
)
Counts = -1 * np.ones(
(ProjectInfo['NumTrajs'], max(ProjectInfo['TrajLengths'])))
print Counts.shape
Savepath = '/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/result/NvOfTrajectory'
plt.figure()
plt.xlabel('Steps')
plt.ylabel('Nv')
plt.hold(False)
for i in range(0, 93):
T = Trajectory.LoadFromHDF(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/sourcedata/Trajectories/trj%d_hc.h5'
% i)
Hcount = hct.count_Helix(T)
plt.title('Nv-steps of Traj%d' % i)
plt.plot(range(len(Hcount)), Hcount, '.')
print 'Save figure to %s/Nvoftraj%d.png' % (Savepath, i)
plt.savefig('%s/Nvoftraj%d.png' % (Savepath, i))
Counts[i, :len(Hcount)] = Hcount[:]
Counts_ma = np.ma.array(Counts, mask=[Counts == -1])
import os, sys
import numpy as np
from msmbuilder import Serializer
from scipy import savetxt, loadtxt
import matplotlib.pyplot as plt
try:
R = Serializer.LoadFromHDF(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/RMSDCluster4.2/Data/RMSD.h5'
)
rmsd = []
for i in range(len(R['Data'])):
for j in range(len(R['Data'][i])):
if R['Data'][i, j] != -1:
rmsd.append(R['Data'][i, j])
except IOError:
print "Can't find RMSD.h5, please run CalculateProjectRMSD.py first to get RMSD.h5."
sys.exit()
try:
Nv = loadtxt(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/result/numhelix_alltraj.txt'
)
nv = []
for i in range(len(Nv)):
for j in range(len(Nv[i])):
if Nv[i, j] != -1:
nv.append(Nv[i, j])
except IOError:
print "Can't find numhelix_alltraj.txt, please run computeNumhelix_alltrajs.py first."
sys.exit()
def get_projectinfo():
projectinfo = Serializer.LoadFromHDF('RMSDCluster4.2/ProjectInfo.h5')
return projectinfo
def get_RMSD():
rmsds = Serializer.LoadFromHDF('RMSDCluster4.2/Data/RMSD-pdb-gen0.h5')
return rmsds['Data']
import sys, os
import numpy as np
import scipy.io
from msmbuilder import arglib
from msmbuilder import Serializer
from msmbuilder import MSMLib
Assignments = Serializer.LoadData("%s" % (sys.argv[1]))
NumStates = max(Assignments.flatten()) + 1
LagTime = sys.argv[2]
Counts = MSMLib.GetCountMatrixFromAssignments(Assignments,
NumStates,
LagTime=LagTime,
Slide=True)
scipy.io.mmwrite('%s' % (sys.argv[3]), Counts)
from scipy import savetxt
from msmbuilder import Serializer
cutoff = 3.0
metrics = 'rmsd'
if metrics.lower() == 'dihedral':
Path = "/Users/tud51931/projects/MSM/msm/ff03-dihedralhybrid/"
metrics = 'Dihedral'
elif metrics.lower() == 'rmsd':
Path = "/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/"
metrics = 'RMSD'
Path = os.path.join(Path, '%sCluster%0.1f' % (metrics, cutoff))
AssignmentFile = os.path.join(Path, "Data", "Assignments.h5")
A = Serializer.LoadFromHDF(AssignmentFile)
StateAssignment = hct.get_StatesAssignments(AssignmentFiles=A)
RMSDFile = os.path.join(Path, "Data", "RMSD.h5")
RMSD = Serializer.LoadFromHDF(RMSDFile)
rmsd_allstates = {}
for state in StateAssignment.keys():
rmsd_singlestate = []
for trajid in StateAssignment[state].keys():
rmsd_singlestate += list(
RMSD['Data'][int(trajid)][StateAssignment[state][trajid]])
rmsd_allstates[int(state)] = rmsd_singlestate
maxstatelength = max([len(i) for i in rmsd_allstates.values()])
StateRMSDs = copy.deepcopy(RMSD)
StateRMSDs['Data'] = -1 * np.ones((len(rmsd_allstates), maxstatelength))
for state in rmsd_allstates.keys():
def save_container(filename, dtype):
s = Serializer({'Data': np.array(minus_ones, dtype=dtype),
'completed_vtrajs': np.zeros((n_vtrajs), dtype=np.bool),
'hashes': hashes})
s.save_to_hdf(filename)
Note: this uses a lag_time of 1 to get the transition counts, and uses
rate estimators that use the *dwell_times*.
The *correct* likelihood function to use for estimating the rate matrix when
the data is sampled at a discrete frequency is open for debate. This
likelihood function doesn't take into account the error in the lifetime estimates
from the discrete sampling. Other methods are currently under development.
Output: tCounts.mtx, K.mtx, Populations.dat, Mapping.dat,
Assignments.Fixed.h5, tCounts.UnSym.mtx""")
parser.add_argument('assignments')
parser.add_argument('symmetrize', choices=['none', 'transpose', 'mle'])
parser.add_argument('outdir')
args = parser.parse_args()
assignments = Serializer.LoadData(args.assignments)
ratemtx_fn = pjoin(args.outdir, 'K.mtx')
tcounts_fn = pjoin(args.outdir, 'tCounts.mtx')
unsym_fn = pjoin(args.outdir, 'tCounts.UnSym.mtx')
mapping_fn = pjoin(args.outdir, 'Mapping.dat')
fixed_fn = pjoin(args.outdir, 'Assignments.Fixed.h5')
pops_fn = pjoin(args.outdir, 'Populations.dat')
if not os.path.exists(args.outdir):
os.mkdir(args.outdir)
outlist = [ratemtx_fn, tcounts_fn, unsym_fn, fixed_fn, pops_fn]
for e in outlist:
arglib.die_if_path_exists(e)
# if lag time is not one, there's going to be a unit mispatch between
# what you get and what you're expecting.
#----------------------------
#October,17,2012
#Guangfeng Zhou
#Dr.Voelz Lab
#Room 100/102, Beury Hall
#Temple University
import sys,os
sys.path.append('/Users/tud51931/scripts/gfzhou')
from msmhcanalysis import SequenceEntropy_states
from msmbuilder import Serializer
import matplotlib.pyplot as plt
import numpy as np
stateentropy = SequenceEntropy_states('HCstrings_states_Dihedral5.2.txt')
rmsdfile = Serializer.LoadFromHDF('StateRMSDs_DihedralCluster5.2.h5')
RMSD = np.ma.array(rmsdfile['Data'],mask=[rmsdfile['Data']==-1])
statermsd = RMSD.mean(1)
plt.figure()
plt.plot(statermsd,stateentropy,'.')
plt.title('StateSequecneEntropy versus StateRMSD')
plt.ylabel('StateSequenceEntropy')
plt.xlabel('StateRMSD(nm)')
plt.savefig('seqentropy_statermsd_dihedralcluster.png')
#plt.show()
def RMSDprediction():
try:
R = Serializer.LoadFromHDF(
'/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter/RMSDCluster4.2/Data/RMSD.h5'
)
except IOError:
print "Can't find RMSD.h5, please run CalculateProjectRMSD.py first to get RMSD.h5."
sys.exit()
RMSD = R['Data']
StatesAsi = hct.get_StatesAssignments(Assignments)
RMSD_states = {}
for state in StatesAsi.keys():
for trajid in StatesAsi[state].keys():
for frame in StatesAsi[state][trajid]:
RMSD_states.setdefault(state, []).append(
RMSD[int(trajid)][int(frame)])
states = [int(i) for i in RMSD_states.keys()]
states.sort()
mean_rmsd_states = []
std_rmsd_states = []
for state in states:
mean_rmsd_states.append(np.mean(RMSD_states['%d' % state]))
std_rmsd_states.append(np.std(RMSD_states['%d' % state]))
#savetxt('mean_numhelix_states0.dat',mean_numhelix_states)
#savetxt('std_numhelix_states0.dat',std_numhelix_states)
print mean_rmsd_states
P0 = np.zeros(len(Population))
for data in Assignments['Data']:
P0[data[0]] += 1
P0 = P0 / P0.sum()
populationslist = []
for k in range(140):
populationslist.append(P0)
P0 *= Tmatrix
RMSD_predicted = np.dot(np.array(populationslist),
np.array(mean_rmsd_states).reshape(-1, 1))
print RMSD_predicted
RMSD_predicted = RMSD_predicted.reshape(1, -1)[0]
plt.figure()
plt.plot(
np.arange(0, 7000, 50),
RMSD_predicted,
'ro',
)
plt.hold(True)
Counts_ma = np.ma.array(RMSD, mask=[RMSD == -1])
RMSD_mean = Counts_ma.mean(0)
RMSD_std = Counts_ma.std(0)
print RMSD_mean
plt.plot(range(len(RMSD_mean)), RMSD_mean, 'b')
plt.title('RMSD-steps')
plt.xlabel('Steps')
plt.ylabel('RMSD')
plt.legend(('RMSD_msm', 'RMSD_rawdata'), loc='upper right')
figname = 'RMSD_prediction_%sCluster%0.1f_tau%d.png' % (metrics, cutoff,
tau)
plt.savefig(figname)
print "Save to %s" % figname
#def GetHCStringsforProject(ProjectInfo)
if __name__ == '__main__':
tau = 50
cutoff = 4.2
metrics = 'rmsd'
if metrics.lower() == 'dihedral':
Path = "/Users/tud51931/projects/MSM/msm/ff03-dihedralhybrid"
metrics = 'Dihedral'
elif metrics.lower() == 'rmsd':
Path = "/Users/tud51931/projects/MSM/msm/ff03-hybridkcenter"
metrics = 'RMSD'
Path = os.path.join(Path, '%sCluster%0.1f' % (metrics, cutoff))
ProjectInfo = Serializer.LoadFromHDF('%s/ProjectInfo.h5' % Path)
Population = loadtxt('%s/lagtime%d/Populations.dat' % (Path, tau))
Assignments = Serializer.LoadFromHDF("%s/lagtime%d/Assignments.Fixed.h5" %
(Path, tau))
Tmatrix = mmread('%s/lagtime%d/tProb.mtx' % (Path, tau))
Gens = '%s/Data/Gens.lh5' % Path
Nvprediction()
EEdistanceprediction()
RMSDprediction()
Rgprediction()
#barchartsforStatesEntropy()
#SequenceEntropy_states()
#GetRgsforGeneratorFile()
#GetHCStringsforTrajectory(Gens)
