pims_nd2 contains a reader for nd2 files produced by NIS Elements. The reader is based on the included SDK from Nikon, which makes it compatible with older versions of nd2 files. The reader is written in the pims framework, enabling easy access to multidimensional files, lazy slicing, and nice display in IPython.
pims_nd2 is implemented on Windows, Linux and OSX systems. To obtain the latest stable version, install via PyPi:
pip install pims_nd2
The ND2 SDK binaries are included in the package and will be copied into the pims_nd2 package folder.
This reader requires pims version 0.3.0.
The following code opens a movie file and displays a frame. Note that frames are only actually read when necessary:
from pims import ND2_Reader
frames = ND2_Reader('some_movie.nd2')
frames[82] # display frame 82
frames.close()
The following code opens the multidimensional demo file included in the package and iterates through the first 3 frames. Note that we use a context manager here. We tell the reader which axis to iterate over and which axes to include in one frame. Also we select which channel to read.
from pims import ND2_Reader
with ND2_Reader('cluster.nd2') as frames:
frames.iter_axes = 't' # 't' is the default already
frames.bundle_axes = 'zyx' # when 'z' is available, this will be default
frames.default_coords['c'] = 1 # 0 is the default setting
for frame in frames[:3]:
# do something with 3D frames in channel 1
The best way to use the reader is using a context manager. If you do not use a context manager, make sure to call frames.close() at the end of your script.
Metadata access can be done on two levels: reader level and frame level.
frames.metadata['mpp'] # calibration in microns per pixel
frames[0].metadata['t_ms'] # time of frame in milliseconds
This reader was developed by Casper van der Wel, as part of his PhD thesis work in Daniela Kraft's group at the Huygens-Kamerlingh-Onnes laboratory, Institute of Physics, Leiden University, The Netherlands. This work was supported by the Netherlands Organisation for Scientific Research (NWO/OCW).