Dynamically generated Python library for reading, writing, and validating Shared Near Infrared Spectroscopy Format (SNIRF) files.
Developed and maintained by the Boston University Neurophotonics Center.
Documentation is generated from source using lazydocs
pip install snirf
pysnirf2 requires Python > 3.6.
pysnirf2 explicitly implements all SNIRF fields so as to provide an extensible object-oriented foundation for SNIRF applications.
Snirf(<path>, <mode>)
opens a SNIRF file at <path>
or creates a new one if it doesn't exist. Use mode 'w' to create a new file, 'r' to read a file, and 'r+' to edit an existing file.
from snirf import Snirf
snirf = Snirf(r'some\path\subj1_run01.snirf', 'r+')
Snirf()
with no arguments creates a temporary file which can be written later using save()
.
snirf = Snirf()
A Snirf
instance wraps a file on disk. It should be closed when you're done reading from it or saving.
snirf.close()
Use a with
statement to ensure that the file is closed when you're done with it:
with Snirf(r'some\path\subj1_run01.snirf', 'r+') as snirf:
# Read/write
snirf.save()
Any Snirf
object can be copied to a new instance in memory, after which the original can be closed.
snirf2 = snirf.copy()
snirf.close()
# snirf2 is free for manipulation
>>> snirf
Snirf at /
filename:
C:\Users\you\some\path\subj1_run01.snirf
formatVersion: v1.0
nirs: <iterable of 2 <class 'pysnirf2.NirsElement'>>
>>> snirf.nirs[0].probe
Probe at /nirs1/probe
correlationTimeDelayWidths: [0.]
correlationTimeDelays: [0.]
detectorLabels: ['D1' 'D2']
detectorPos2D: [[30. 0.]
[ 0. 30.]]
detectorPos3D: [[30. 0. 0.]
[ 0. 30. 0.]]
filename:
C:\Users\you\some\path\subj1_run01.snirf
frequencies: [1.]
landmarkLabels: None
landmarkPos2D: None
landmarkPos3D: None
location: /nirs/probe
momentOrders: None
sourceLabels: ['S1']
sourcePos2D: [[0. 0.]]
sourcePos3D: [[0.]
[0.]
[0.]]
timeDelayWidths: [0.]
timeDelays: [0.]
useLocalIndex: None
wavelengths: [690. 830.]
wavelengthsEmission: None
Assign a new value to a field
>>> snirf.nirs[0].metaDataTags.SubjectID = 'subj1'
>>> snirf.nirs[0].metaDataTags.SubjectID
'subj1'
>>> snirf.nirs[0].probe.detectorPos3D[0, :] = [90, 90, 90]
>>> snirf.nirs[0].probe.detectorPos3D
array([[90., 90., 90.],
[ 0., 30., 0.]])
Note: assignment via slicing is not possible in
dynamic_loading
mode.
Indexed groups are defined by the SNIRF file format as groups of the same type which are indexed via their name + a 1-based index, i.e. data1
, data2
, ... or stim1
, stim2
, stim3
, ...
pysnirf2 provides an iterable interface for these groups using Pythonic 0-based indexing, i.e. data[0]
, data[1]
, ... or stim[0]
, stim[1]]
, stim[2]
, ...
>>> snirf.nirs[0].stim
<iterable of 0 <class 'pysnirf2.StimElement'>>
>>> len(nirs[0].stim)
0
To add an indexed group, use the appendGroup()
method of any IndexedGroup
class. Indexed groups are created automatically. nirs
is an indexed group.
>>> snirf.nirs[0].stim.appendGroup()
>>> len(nirs[0].stim)
1
>>> snirf.nirs[0].stim[0]
StimElement at /nirs/stim2
data: None
dataLabels: None
filename:
C:\Users\you\some\path\subj1_run01.snirf
name: None
To remove an indexed group
del snirf.nirs[0].stim[0]
Overwrite the open file
snirf.save()
Save As in a new location
snirf.save(r'some\new\path\subj1_run01_edited.snirf')
The save()
function can be called for any group or indexed group:
snirf.nirs[0].metaDataTags.save('subj1_run01_edited_metadata_only.snirf')
For larger files, it may be useful to load data dynamically: data will only be loaded on access, and only changed datasets will be written on save()
. When creating a new Snirf
instance, set dynamic_loading
to True
(Default False
).
snirf = Snirf(r'some\path\subj1_run01.snirf', 'r+', dynamic_loading=True)
Note: in dynamic loading mode, array data cannot be modified with indices like in the example above:
>>> snirf = Snirf(TESTPATH, 'r+', dynamic_loading=True) >>> snirf.nirs[0].probe.detectorPos3D array([[30., 0., 0.], [ 0., 30., 0.]]) >>> snirf.nirs[0].probe.detectorPos3D[0, :] = [90, 90, 90] >>> snirf.nirs[0].probe.detectorPos3D array([[30., 0., 0.], [ 0., 30., 0.]])To modify an array in
dynamic_loading
mode, assign it, modify it, and assign it back to the Snirf object.>>> detectorPos3D = snirf.nirs[0].probe.detectorPos3D >>> detectorPos3D[0, :] = [90, 90, 90] >>> snirf.nirs[0].probe.detectorPos3D = detectorPos3D array([[90., 90., 90.], [ 0., 30., 0.]])
pysnirf2 features functions for validating SNIRF files against the specification and generating detailed error reports.
result = snirf.validate()
To validate a SNIRF file on disk
from snirf import validateSnirf
result = validateSnirf(r'some\path\subj1_run01.snirf')
assert result, 'Invalid SNIRF file!\n' + result.display() # Crash and display issues if the file is invalid.
The validation functions return a ValidationResult
instance which contains details about the SNIRF file.
To view the validation result:
>>> result.display(severity=3) # Display all fatal errors
<pysnirf2.pysnirf2.ValidationResult object at 0x000001C0CCF05A00>
/nirs1/data1/measurementList103/dataType FATAL REQUIRED_DATASET_MISSING
/nirs1/data1/measurementList103/dataTypeIndex FATAL REQUIRED_DATASET_MISSING
/nirs1/data1 FATAL INVALID_MEASUREMENTLIST
Found 668 OK (hidden)
Found 635 INFO (hidden)
Found 204 WARNING (hidden)
Found 3 FATAL
File is INVALID
To look at a particular result:
>>> result.errors[2]
<pysnirf2.pysnirf2.ValidationIssue object at 0x000001C0CB502F70>
location: /nirs1/data1
severity: 3 FATAL
name: 8 INVALID_MEASUREMENTLIST
message: The number of measurementList elements does not match the second dimension of dataTimeSeries
The full list of validation results result.issues
can be explored programatically.
The interface and validator are generated via metacode that downloads and parses the latest SNIRF specification.
See \gen for details.