wrapper scripts for SOFIMA to batch jobs to the janelia cluster.
the entry point is 2d-align.sh, which submits two jobs, the first (2d-flow-mesh.py) handles the GPU intensive portions of the algorithm, and a second dependent one (2d-invmap.py) which uses only the CPU and also a lot of RAM. data loading and saving code is in a plugin (see data-*.py).
a similar workflow exists for 3d data, which aligns the overlapping portions of two volumes instead of just two adjacent slices.
conda create -n multi-sem
conda install jax tensorstore
conda install matplotlib # optional
pip install git+https://github.com/google-research/sofima
sofima also installs numpy and connectomics which these scripts directly use
matplotlib is needed only for testing purposes. see {2,3}d-test.py and the development section below.
manually edit 2d-align.sh to set basepath to the location of this
repository, and the arguments to bsub to the number of slots required, etc.
manually edit one of the data-.py (e.g. to set zbase to the location of
your data and itop and ibot to the two slices you want to align).
alternatively, write your own data-loader plugin defining the same functions
that are in data-.py.
then execute: ./2d-align.sh <data-loader> <patch-size> <stride> <batch-size>
and are in units of pixels and set the XY spatial context used for flow field estimation and the XY distance between centers of adjacent patches, respectively.
set such that all GPU cores are used without exceeding GPU RAM
designed to only align the bottom of one 3D block to the top of another
uses the zarr folder of the input data as scratch space for intermediate results
flow fields are only calculated at a single resolution
unit tests are in {2,3}d-test.py and use data-test-{2,3}d.py. for 2d-test.py the following figure should be generated:
