|
Rosetta
|
Functions | |
| def | find_nonmodulo_autocorrelation (x) |
| def | find_autocorrelation (x) |
Variables | |
| parser = helpers.define_args('torsion', 'plot') | |
| type | |
| float | |
| default | |
| arguments = parser.parse_args() | |
| iterations = helpers.load_array(arguments.job, 'iterations') | |
| torsions = helpers.load_arrays(arguments.job, 'torsions') | |
| def | autocorrelation = find_autocorrelation(torsions[key]) |
| color = pylab.cm.Dark2(numpy.median(autocorrelation)) | |
| label | |
| key | |
| def autocorrelation.find_autocorrelation | ( | x | ) |
This homebrew autocorrelation function is specifically written to handle modulo variable like dihedral angles. Think of the cosine function as an alternative, and more appropriate, metric for overlap.
References helpers.connect_to_database(), ObjexxFCL.len(), numeric::conversions.radians(), range, and sum().
| def autocorrelation.find_nonmodulo_autocorrelation | ( | x | ) |
The standard correlation function isn't meant for modulo variables, like dihedral angles. I'm sure this version is much faster, but it will choke on angles constantly jumping between 180 and -180.
References ObjexxFCL.len().
| autocorrelation.arguments = parser.parse_args() |
| autocorrelation.autocorrelation = find_autocorrelation(torsions[key]) |
| autocorrelation.color = pylab.cm.Dark2(numpy.median(autocorrelation)) |
| autocorrelation.default |
| autocorrelation.float |
| autocorrelation.iterations = helpers.load_array(arguments.job, 'iterations') |
| autocorrelation.key |
| autocorrelation.label |
| autocorrelation.parser = helpers.define_args('torsion', 'plot') |
| autocorrelation.torsions = helpers.load_arrays(arguments.job, 'torsions') |
| autocorrelation.type |