def __init__(self,
clustercenters,
metric='euclidean',
stride=1,
n_jobs=None,
skip=0):
super(AssignCenters, self).__init__(metric=metric, n_jobs=n_jobs)
if isinstance(clustercenters, str):
from pyemma.coordinates.data import create_file_reader
reader = create_file_reader(clustercenters, None, None)
clustercenters = reader.get_output()[0]
else:
clustercenters = np.array(clustercenters,
dtype=np.float32,
order='C')
# sanity check.
if not clustercenters.ndim == 2:
raise ValueError('cluster centers have to be 2d')
self.set_params(clustercenters=clustercenters,
metric=metric,
stride=stride,
skip=skip)
# since we provided centers, no estimation is required.
self._estimated = True
def __init__(self, clustercenters, metric='euclidean', stride=1, n_jobs=None, skip=0):
r"""Assigns given (pre-calculated) cluster centers. If you already have
cluster centers from somewhere, you use this class to assign your data to it.
.. deprecated:: 2.5.11
Use the deeptime
`ClusterModel <https://deeptime-ml.github.io/latest/api/generated/deeptime.clustering.ClusterModel.html>`__
and its `transform` method instead. Will be removed in PyEMMA 3.
Parameters
----------
clustercenters : path to file (csv) or npyfile or ndarray
cluster centers to use in assignment of data
metric : str
metric to use during clustering ('euclidean', 'minRMSD')
stride : int
stride
n_jobs : int or None, default None
Number of threads to use during assignment of the data.
If None, all available CPUs will be used.
skip : int, default=0
skip the first initial n frames per trajectory.
Examples
--------
Assuming you have stored your centers in a CSV file:
>>> from pyemma.coordinates.clustering import AssignCenters
>>> from pyemma.coordinates import pipeline
>>> reader = ... # doctest: +SKIP
>>> assign = AssignCenters('my_centers.dat') # doctest: +SKIP
>>> disc = pipeline(reader, cluster=assign) # doctest: +SKIP
>>> disc.parametrize() # doctest: +SKIP
"""
super(AssignCenters, self).__init__(metric=metric, n_jobs=n_jobs)
if isinstance(clustercenters, str):
from pyemma.coordinates.data import create_file_reader
reader = create_file_reader(clustercenters, None, None)
clustercenters = reader.get_output()[0]
else:
clustercenters = np.array(clustercenters, dtype=np.float32, order='C')
# sanity check.
if not clustercenters.ndim == 2:
raise ValueError('cluster centers have to be 2d')
self.set_params(clustercenters=clustercenters, metric=metric, stride=stride, skip=skip)
# since we provided centers, no estimation is required.
self._estimated = True
def __init__(self, clustercenters, metric='euclidean'):
super(AssignCenters, self).__init__(metric=metric)
if isinstance(clustercenters, six.string_types):
from pyemma.coordinates.data import create_file_reader
reader = create_file_reader(clustercenters, None, None)
self._clustercenters = reader.get_output()[0]
else:
self._clustercenters = np.array(clustercenters,
dtype=np.float32,
order='C')
# sanity check.
if not self.clustercenters.ndim == 2:
raise ValueError('cluster centers have to be 2d')
# since we provided centers, this transformer is already parametrized.
self._parametrized = True