tools for:
- handling data from Baylor and Fiji Image stacks
- performing a 3D alignment solve between landmarks
- transforming points based on a 3D solve result
- create neuroglancer links for checking visualization
We are not currently supporting this code, but simply releasing it to the community AS IS but are not able to provide any guarantees of support, as it is under active development. The community is welcome to submit issues, but you should not expect an active response.
Supported by the Intelligence Advanced Research Projects Activity (IARPA) via Department of Interior / Interior Business Center (DoI/IBC) contract number D16PC00004. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright annotation thereon. Disclaimer: The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of IARPA, DoI/IBC, or the U.S. Government.
- data/staged_transform_args.json : input arguments for solve 2P->EM
- data/inverse_staged_transform_args.json : the other way
- data/staged_transform_solution.json : transform solution 2P -> EM
- data/inverse_staged_transform_solution.json : the other way
- data/17797_2Pfix_EMmoving_20191010_1652_piecewise_trial_updated_Masterleave_outs.json : leave-one-out residuals
- data/17797_2Pfix_EMmoving_20191010_1652_piecewise_trial_updated_Masterinverse_leave_outs.json : the other way
import json
import coregister.solve as cs
with open("./data/staged_transform_args.json", "r") as f:
j =json.load(f)
s = cs.Solve3D(input_data=j, args=['--output_json', 'solved_transform_out.json'])
s.run()
from coregister.transform.transform import Transform
with open('./solved_transform_out.json', 'r') as f:
tfj = json.load(f)
t = Transform(json=tfj)
src = s.data['src'][0:5]
dst = s.data['dst'][0:5]
src
Out[20]:
array([[0.804354, 0.387835, 0.274648],
[0.869191, 0.138524, 0.156319],
[0.85027 , 0.257353, 0.067352],
[0.924324, 0.750244, 0.225444],
[0.934366, 0.182291, 0.188485]])
dst
Out[21]:
array([[1172669.4, 717762.7, 282148.8],
[1391713. , 593754.5, 337574.1],
[1287230.6, 508049.5, 321659.2],
[ 869550.9, 684943.2, 399641.7],
[1363565.2, 611139.6, 398796.7]])
t.tform(src)
Out[22]:
array([[1172669.3087167 , 717762.76429371, 282148.60118503],
[1391715.11237805, 593753.07429628, 337574.56376036],
[1287230.51081691, 508049.13923649, 321659.30456869],
[ 869551.3786766 , 684942.05607145, 399640.98195057],
[1363565.80927674, 611138.89843287, 398797.29234907]])from the root dir of this repo. If running as an installed package, you'll need to copy this data directory somewhere with r/w permissions. :: python fit_and_predict.py
average residual [dst units]: 3.1620 worst points Pt-3044 0.020422 Pt-727 0.035287 Pt-1047 0.042640 Pt-2182 0.047620 Pt-236 0.047632 Pt-2211 0.049158 Pt-2757 0.053443 Pt-1862 0.057763 Pt-768 0.058496 Pt-25103 0.058756 PolynomialModel residual 9932.802718 ChunkedModel residual 7296.263690 ChunkedModel residual 5909.392451 SplineModel residual 4503.411981 27 cntrls moved 22585.214699 SplineModel residual 3781.220908 125 cntrls moved 18947.336478 SplineModel residual 3173.745507 1000 cntrls moved 6918.062524 SplineModel residual 2893.618486 1728 cntrls moved 3360.592100 SplineModel residual 3.161994 2936 cntrls moved 2892.010575 wrote data/17797_2Pfix_EMmoving_20191010_1652_piecewise_trial_updated_Master_updated.csv
this just performed a staged solve, showing residuals and control point motions for the specified transform steps. Refer to fit_and_predict.py for more details.
Running this can be time-consuming: :: python leave_one_out.py
For testing, one can change the leave-out fraction inside the file to something smaller than 1 (for example 0.002 will run just a few). I tend to run it on a cluster node. See coreg.pbs.