Merge pull request #276 from rmcgibbo/trim

Fix #252 on master

Author: rmcgibbo

scripts/CalculateImpliedTimescales.py

@@ -17,14 +17,39 @@
 # along with this program; if not, write to the Free Software
 # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 
+import logging
 import numpy as np
-
 from msmbuilder import arglib
 from msmbuilder import msm_analysis
-
-import logging
 logger = logging.getLogger('msmbuilder.scripts.CalculateImpliedTimescales')
 
+parser = arglib.ArgumentParser(description="""
+\nCalculates the implied timescales of a set of assigned data, up to
+the argument 'lagtime'. Returns: ImpliedTimescales.dat, a flat file that
+contains all the lag times.\n""")
+parser.add_argument('assignments', type=str)
+parser.add_argument('lagtime', help="""The lagtime range to calculate.
+    Pass two ints as X,Y with NO WHITESPACE, where X is the lowest
+    timescale you want and Y is the biggest. EG: '-l 5,50'.""")
+parser.add_argument('output', help="""The name of the  implied
+    timescales data file (use .dat extension)""", default='ImpliedTimescales.dat')
+parser.add_argument('procs', help='''Number of concurrent processes
+    (cores) to use''', default=1, type=int)
+parser.add_argument('eigvals', help="""Number of slowest implied timescales to
+    retrieve at each lag time. Note: an n-state model will have n-1
+    implied timescales.""", default=10, type=int)
+parser.add_argument('interval', help="""Interval between times (intervals)
+    to calculate lagtimes for""", default=1, type=int)
+parser.add_argument('symmetrize', help="""Method by which to estimate a
+    symmetric counts matrix. Symmetrization ensures reversibility, but may skew
+    dynamics. We recommend maximum likelihood estimation (MLE) when tractable,
+    else try Transpose. It is strongly recommended you read the documentation
+    surrounding this choice.""", default='MLE',
+                    choices=['MLE', 'Transpose', 'None'])
+parser.add_argument('notrim', help="""Do not apply an ergodic trim.
+    By default, we keep only the largest observed ergodic subset of the data.
+    supplied, keeps everything.""", action='store_true')
+
 
 def run(MinLagtime, MaxLagtime, Interval, NumEigen, AssignmentsFn, trimming,
         symmetrize, nProc):
@@ -34,40 +59,13 @@ def run(MinLagtime, MaxLagtime, Interval, NumEigen, AssignmentsFn, trimming,
     logger.info("Building MSMs at the following lag times: %s", lagTimes)
 
     # Get the implied timescales (eigenvalues)
-    impTimes = msm_analysis.get_implied_timescales(AssignmentsFn, lagTimes,
-        n_implied_times=NumEigen, sliding_window=True, trimming=trimming,
-        symmetrize=symmetrize, n_procs=nProc)
-
+    impTimes = msm_analysis.get_implied_timescales(
+        AssignmentsFn, lagTimes, n_implied_times=NumEigen, sliding_window=True,
+        trimming=trimming, symmetrize=symmetrize, n_procs=nProc)
     return impTimes
 
 
 if __name__ == "__main__":
-    parser = arglib.ArgumentParser(description="""
-\nCalculates the implied timescales of a set of assigned data, up to
-the argument 'lagtime'. Returns: ImpliedTimescales.dat, a flat file that
-contains all the lag times.\n""")
-    parser.add_argument('assignments', type=str)
-    parser.add_argument('lagtime', help="""The lagtime range to calculate.
-        Pass two ints as X,Y with NO WHITESPACE, where X is the lowest
-        timescale you want and Y is the biggest. EG: '-l 5,50'.""")
-    parser.add_argument('output', help="""The name of the  implied
-        timescales data file (use .dat extension)""", default='ImpliedTimescales.dat')
-    parser.add_argument('procs', help='''Number of concurrent processes
-        (cores) to use''', default=1, type=int)
-    parser.add_argument('eigvals', help="""Number of slowest implied timescales to
-        retrieve at each lag time. Note: an n-state model will have n-1
-        implied timescales.""", default=10, type=int)
-    parser.add_argument('interval', help="""Interval between times (intervals)
-        to calculate lagtimes for""", default=1, type=int)
-    parser.add_argument('symmetrize', help="""Method by which to estimate a
-        symmetric counts matrix. Symmetrization ensures reversibility, but may skew
-        dynamics. We recommend maximum likelihood estimation (MLE) when tractable,
-        else try Transpose. It is strongly recommended you read the documentation
-        surrounding this choice.""", default='MLE',
-        choices=['MLE', 'Transpose', 'None'])
-    parser.add_argument('trim', help="""Whether or not to apply an ergodic trim.
-        If true, keeps only the largest observed ergodic subset of the data, if
-        false, keeps everything. Default: True.""", default=True, type=bool)
     args = parser.parse_args()
     arglib.die_if_path_exists(args.output)
 
@@ -79,7 +77,8 @@ def run(MinLagtime, MaxLagtime, Interval, NumEigen, AssignmentsFn, trimming,
     if args.symmetrize in ["None", "none", None]:
         args.symmetrize = None
 
-    impTimes = run(MinLagtime, MaxLagtime, args.interval, args.eigvals, args.assignments,
-        args.trim, args.symmetrize, args.procs)
+    impTimes = run(
+        MinLagtime, MaxLagtime, args.interval, args.eigvals, args.assignments,
+        (not args.notrim), args.symmetrize, args.procs)
     np.savetxt(args.output, impTimes)
     logger.info("Saved output to %s", args.output)