# This script is part of navis (http://www.github.com/navis-org/navis).
# Copyright (C) 2018 Philipp Schlegel
#
# This program is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
"""This module contains functions to read R data (.rda) files."""
import rdata
import warnings
import numpy as np
import pandas as pd
from typing import Any, Mapping, Union
from .. import config, utils, core
__all__ = ['read_rda']
# Set up logging
logger = config.get_logger(__name__)
[docs]
def read_rda(f: str,
combine: bool = True,
neurons_only: bool = True,
**kwargs) -> 'core.NeuronList':
"""Read objects from nat R data (.rda) file.
Currently supports parsing neurons, dotprops and mesh3d. Note that this is
rather slow and I do not recommend doing this for large collections of
neurons. For large scale conversion I recommend using the R interface
(``navis.interfaces.r``, see online tutorials) via ``rpy2``.
Parameters
----------
f : str
Filepath.
combined : bool
What to do if there are multiple neuronlists contained
in the RDA files. By default, we will combine them into
a single NeuronList but you can also choose to keep them
as separate neuronlists.
neurons_only : bool
Whether to only parse and return neurons and dotprops
found in the RDA file.
**kwargs
Keyword arguments passed to the construction of
`Tree/MeshNeuron/Dotprops`. You can use this to e.g. set
meta data.
Returns
-------
navis.NeuronList
If ``combine=True`` and ``neurons_only=True`` returns
a single NeuronList with the parsed neurons.
dict
If ``combine=False`` or ``neurons_only=False`` returns
a dictionary with the original R object name as key and
the parsed object as value.
"""
# Parse the file
parsed = rdata.parser.parse_file(f)
# Now convert to Python objects
with warnings.catch_warnings():
warnings.simplefilter("ignore")
converted = rdata.conversion.convert(parsed, CLASS_MAP_EXT)
# Some clean-up
for k, v in converted.items():
# Convert single neurons to neuronlist
if isinstance(v, core.BaseNeuron):
converted[k] = core.NeuronList(v)
# Give volumes a name
elif isinstance(v, core.Volume):
converted[k].name = k
if combine:
nl = core.NeuronList([n for n in converted.values() if isinstance(n, core.NeuronList)])
if nl:
converted = {k: v for k, v in converted.items() if not isinstance(v, core.NeuronList)}
converted['neurons'] = nl
if neurons_only:
if combine:
converted = converted['neurons']
else:
converted = {k: v for k, v in converted.items() if isinstance(v, core.NeuronList)}
return converted
def neuronlist_constructor(obj: Any,
attrs: Mapping[Union[str, bytes], Any],
) -> 'core.NeuronList':
"""Convert nat neuronlists to navis NeuronLists."""
# Set IDs
neurons = []
for k, n in obj.items():
if isinstance(n, (core.BaseNeuron, core.NeuronList)):
n.id = k
neurons.append(n)
else:
logger.warning(f'Unexpected object in neuronlist: {type(n)}. '
'Possible parsing error.')
# Turn into NeuronList
nl = core.NeuronList(neurons)
# Now parse extra attributes DataFrame
df = attrs.get('df', None)
if isinstance(df, pd.DataFrame):
# Make sure we have still the correct order
nl = nl.idx[attrs['names']]
for col in df:
# Skip non-string columns
if not isinstance(col, str):
continue
# Skip some columns
if col.lower() in ['type', 'idx']:
continue
if col.lower() in nl[0].__dict__.keys():
continue
for n, v in zip(nl, df[col].values):
# Register
n._register_attr(col.lower(), v)
return nl
def dotprops_constructor(obj: Any,
attrs: Mapping[Union[str, bytes], Any],
) -> 'core.Dotprops':
"""Convert nat dotprops to navis Dotprops."""
pts = np.asarray(obj.pop('points'))
vect = np.asarray(obj.pop('vect'))
alpha = np.asarray(obj.pop('alpha'))
k = int(attrs.get('k', 1)[0])
file = attrs.get('file', [None])[0]
return core.Dotprops(points=pts, k=k, alpha=alpha, vect=vect, file=file)
def volume_constructor(obj: Any,
attrs: Mapping[Union[str, bytes], Any],
) -> 'core.Volume':
"""Convert e.g. mesh3d to navis Volume."""
if 'vb' in obj and 'it' in obj:
verts = np.asarray(obj.pop('vb'))[:3, :].T
faces = np.asarray(obj.pop('it')).T - 1
return core.Volume(vertices=verts, faces=faces)
elif 'Vertices' in obj and "Regions" in obj:
verts = obj['Vertices'][['X', 'Y', 'Z']].values
# If only one region
if len(obj['Regions']) == 1:
region = list(obj['Regions'].keys())[0]
faces = obj['Regions'][region][['V1', 'V2', 'V3']].values - 1
return core.Volume(vertices=verts, faces=faces)
else:
volumes = []
for r in obj['Regions']:
faces = obj['Regions'][r][['V1', 'V2', 'V3']].values - 1
volumes.append(core.Volume(vertices=verts, faces=faces, name=r))
return volumes
else:
logger.warning('Unable to construct Volume from R object of type '
f'"{attrs["class"]}". Returning raw data')
return obj
def neuron_constructor(obj: Any,
attrs: Mapping[Union[str, bytes], Any],
) -> 'core.TreeNeuron':
"""Convert nat neuron/catmaidneuron to navis TreeNeuron."""
# Data to skip
DO_NOT_USE = ['nTrees', 'SegList', 'NumPoints', 'StartPoint', 'EndPoints',
'BranchPoints', 'NumSegs']
# Construct neuron from just the nodes
n = core.TreeNeuron(obj.pop('d'))
# R uses diameter, not radius - let's fix that
if 'radius' in n.nodes.columns:
has_rad = n.nodes.radius.fillna(0) > 0
n.nodes.loc[has_rad, 'radius'] = n.nodes.loc[has_rad, 'radius'] / 2
# If this is a CATMAID neuron, we assume it's in nanometers
if 'catmaidneuron' in attrs.get('class', []):
n.units = 'nm'
# Reuse ID
if 'skid' in obj:
skid = obj.pop('skid')
if utils.is_iterable(skid):
n.id = skid[0]
else:
n.id = skid
# Try attaching other data
for k, v in obj.items():
if k in DO_NOT_USE:
continue
try:
setattr(n, k, v)
except BaseException:
pass
return n
CLASS_MAP_EXT = {**rdata.conversion.DEFAULT_CLASS_MAP,
"neuronlist": neuronlist_constructor,
"neuron": neuron_constructor,
"mesh3d": volume_constructor,
"hxsurf": volume_constructor,
"dotprops": dotprops_constructor}