Skip to content

SpikeTrainList improvements #1112

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
JuliaSprenger merged 3 commits into from
May 16, 2022
Merged

SpikeTrainList improvements #1112

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
3e3acf9
fix a bug when annotating SpikeTrainList with strings
apdavison Apr 15, 2022
20a8c5c
add properties `t_stop`, `t_start` and `all_channel_ids` to SpikeTrai…
apdavison Apr 15, 2022
11c7963
try to make pep8speaks happy
apdavison Apr 15, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
84 changes: 78 additions & 6 deletions neo/core/spiketrainlist.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6,6 +6,7 @@
neuron/channel the spike is from).
"""

import warnings
import numpy as np
import quantities as pq
from .spiketrain import SpikeTrain, normalize_times_array
Expand All @@ -15,6 +16,22 @@ def is_spiketrain_or_proxy(obj):
return isinstance(obj, SpikeTrain) or getattr(obj, "proxy_for", None) == SpikeTrain


def unique(quantities):
"""np.unique doesn't work with a list of quantities of different scale,
this function can be used instead."""
# todo: add a tolerance to handle floating point discrepancies
# due to scaling.
if len(quantities) > 0:
common_units = quantities[0].units
scaled_quantities = pq.Quantity(
[q.rescale(common_units) for q in quantities],
common_units)
return np.unique(scaled_quantities)
else:
return quantities



class SpikeTrainList(object):
"""
This class contains multiple spike trains, and can represent them
Expand Down Expand Up @@ -70,7 +87,7 @@ def __init__(self, items=None, segment=None):
self._spike_time_array = None
self._channel_id_array = None
self._all_channel_ids = None
self._spiketrain_metadata = None
self._spiketrain_metadata = {}
self.segment = segment

def __iter__(self):
Expand Down Expand Up @@ -263,7 +280,10 @@ def from_spike_time_array(cls, spike_time_array, channel_id_array,
"t_stop": t_stop
}
for name, ann_value in annotations.items():
if len(ann_value) != len(obj):
if (not isinstance(ann_value, str)
and hasattr(ann_value, "__len__")
and len(ann_value) != len(all_channel_ids)
):
raise ValueError(f"incorrect length for annotation '{name}'")
obj._annotations = annotations
return obj
Expand All @@ -278,10 +298,14 @@ def _spiketrains_from_array(self):
mask = self._channel_id_array == channel_id
times = self._spike_time_array[mask]
spiketrain = SpikeTrain(times, **self._spiketrain_metadata)
spiketrain.annotations = {
name: value[i]
for name, value in self._annotations.items()
}
for name, value in self._annotations.items():
if (not isinstance(value, str)
and hasattr(value, "__len__")
and len(value) == len(self._all_channel_ids)
):
spiketrain.annotate(**{name: value[i]})
else:
spiketrain.annotate(**{name: value})
spiketrain.annotate(channel_id=channel_id)
spiketrain.segment = self.segment
self._items.append(spiketrain)
Expand Down Expand Up @@ -319,3 +343,51 @@ def multiplexed(self):
self._spike_time_array = np.hstack(spike_times) * self._items[0].units
self._channel_id_array = np.hstack(channel_ids)
return self._channel_id_array, self._spike_time_array

@property
def t_start(self):
if "t_start" in self._spiketrain_metadata:
return self._spiketrain_metadata["t_start"]
else:
t_start_values = unique([item.t_start for item in self._items
if isinstance(item, SpikeTrain)]) # ignore proxy objects
if len(t_start_values) == 0:
raise ValueError("t_start not defined for an empty spike train list")
elif len(t_start_values) > 1:
warnings.warn("Found multiple values of t_start, returning the earliest")
t_start = t_start_values.min()
else:
t_start = t_start_values[0]
self._spiketrain_metadata["t_start"] = t_start
return t_start

@property
def t_stop(self):
if "t_stop" in self._spiketrain_metadata:
return self._spiketrain_metadata["t_stop"]
else:
t_stop_values = unique([item.t_stop for item in self._items
if isinstance(item, SpikeTrain)]) # ignore proxy objects
if len(t_stop_values) == 0:
raise ValueError("t_stop not defined for an empty spike train list")
elif len(t_stop_values) > 1:
warnings.warn("Found multiple values of t_stop, returning the latest")
t_stop = t_stop_values.max()
else:
t_stop = t_stop_values[0]
self._spiketrain_metadata["t_stop"] = t_stop
return t_stop

@property
def all_channel_ids(self):
if self._all_channel_ids is None:
self._all_channel_ids = []
for i, spiketrain in enumerate(self._items):
if ("channel_id" in spiketrain.annotations
and isinstance(spiketrain.annotations["channel_id"], int)
):
ch_id = spiketrain.annotations["channel_id"]
else:
ch_id = i
self._all_channel_ids.append(ch_id)
return self._all_channel_ids
16 changes: 15 additions & 1 deletion neo/test/coretest/test_spiketrainlist.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -69,7 +69,9 @@ def setUp(self):
units='ms',
t_start=0 * pq.ms,
t_stop=100.0 * pq.ms,
identifier=["A", "B", "C", "D"] # annotation
identifier=["A", "B", "C", "D"], # separate annotation for each SpikeTrain
global_str="some string annotation", # global annotations, same for each SpikeTrain
global_int=42
)

self.stl_from_obj_list = SpikeTrainList(items=(
Expand Down Expand Up @@ -106,6 +108,13 @@ def test_create_from_spiketrain_array(self):
self.assertEqual(as_list[1].annotations["identifier"], "B")
self.assertEqual(as_list[2].annotations["identifier"], "C")
self.assertEqual(as_list[3].annotations["identifier"], "D")
self.assertEqual(as_list[0].annotations["global_str"], "some string annotation")
self.assertEqual(as_list[3].annotations["global_str"], "some string annotation")
self.assertEqual(as_list[2].annotations["global_int"], 42)
self.assertEqual(as_list[1].annotations["global_int"], 42)
self.assertEqual(self.stl_from_array.t_stop, 100.0 * pq.ms)
self.assertEqual(self.stl_from_array.all_channel_ids, (0, 1, 2, 3))


def test_create_from_spiketrain_list(self):
as_list = list(self.stl_from_obj_list)
Expand All @@ -117,6 +126,9 @@ def test_create_from_spiketrain_list(self):
np.array([1.1, 88.5]))
assert_array_equal(as_list[3].times.rescale(pq.ms).magnitude,
np.array([]))
self.assertAlmostEqual(self.stl_from_obj_list.t_stop, 100.0 * pq.ms)
self.assertEqual(self.stl_from_obj_list.all_channel_ids, [101, 102, 103, 104])


def test_create_from_spiketrain_list_incl_proxy(self):
as_list = list(self.stl_from_obj_list_incl_proxy)
Expand All @@ -127,6 +139,8 @@ def test_create_from_spiketrain_list_incl_proxy(self):
assert isinstance(as_list[2], SpikeTrainProxy)
assert_array_equal(as_list[3].times.rescale(pq.ms).magnitude,
np.array([]))
self.assertAlmostEqual(self.stl_from_obj_list_incl_proxy.t_stop, 100.0 * pq.ms)
self.assertEqual(self.stl_from_obj_list_incl_proxy.all_channel_ids, [0, 1, 2, 3])

def test_str(self):
target = "SpikeTrainList containing 8 spikes from 4 neurons"
Expand Down