Skip to content
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.

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

CNMF #5

Closed
whusym opened this issue Feb 23, 2018 · 6 comments
Closed

CNMF #5

whusym opened this issue Feb 23, 2018 · 6 comments
Assignees
Labels
enhancement New feature or request

Comments

@whusym
Copy link
Contributor

whusym commented Feb 23, 2018

We can use CNMF to improve the performance of NMF

@whusym whusym created this issue from a note in Neuron Finding (To Do) Feb 23, 2018
@whusym whusym moved this from To Do to In Progress in Neuron Finding Feb 23, 2018
@whusym whusym added the enhancement New feature or request label Feb 23, 2018
@melanieihuei
Copy link
Contributor

CaImAn works fine but only in the specific directory
working on the reproduced-CNMF

@melanieihuei
Copy link
Contributor

(environment update)

  • add future package to environment.txt to make past.utils works
  • do not install caiman directly from their github in case it covers those unavailable packages
  • new environment set up:
    pip install -r requirements.txt
    python setup.y install
  • CNMFWRAP from thunder-extraction is based on CNMF from CaImAn, make sure it fits python 3.5 but not python 2.7 (map() generates iterator in py35 and reports error when importing cnmf)
  • change all caiman imports to their actual path

@melanieihuei
Copy link
Contributor

(running errors update)

  • running CNMF from reproduced-CaImAn under caiman-env:
    errors:
  File "example_cnmf.py", line 21, in <module>
    model,temporaldata = algorithm.fit(data)
TypeError: 'CNMF' object is not iterable
  • running CNMFWRAP from thunder-extraction under caiman-env:
    errors:
Traceback (most recent call last):
  File "example_cnmfwrap.py", line 13, in <module>
    model = algorithm.fit(data, chunk_size=(100,200))
  File "/Users/melanie/csci8360/p3/neurofinder.00.00/extraction/algorithms/cnmfwrap.py", line 28, in fit
    sources = asarray(blocks.map_generic(self._get))
  File "/Users/melanie/anaconda3/envs/caiman/lib/python3.5/site-packages/thunder/blocks/blocks.py", line 63, in map_generic
    return self.values.map_generic(func)[0]
  File "/Users/melanie/anaconda3/envs/caiman/lib/python3.5/site-packages/thunder/blocks/local.py", line 117, in map_generic
    arr[i] = func(x)
  File "/Users/melanie/csci8360/p3/neurofinder.00.00/extraction/algorithms/cnmfwrap.py", line 49, in _get
    model, temporaldata = algorithm.fit(block)
TypeError: 'CNMF' object is not iterable
  • running CNMFWRAP from thunder-extraction under caiman-env` of neuron-example:
    errors:
RuntimeWarning: invalid value encountered in maximum
  cc = np.maximum(YrA[:, ii] + Cin[ii], -np.Inf)

@melanieihuei
Copy link
Contributor

just a note for myself, im creating new environment and want to avoid those nonsense problems...

@melanieihuei
Copy link
Contributor

able to run the script without having the whole repo of CaImAn (required to follow the environment setting in readme.)
need to change the neuron pixels location to neuron coordinates

@melanieihuei
Copy link
Contributor

Successfully transform pixels into coordinates as required format. Implement to all the testing sets.

@melanieihuei melanieihuei mentioned this issue Mar 6, 2018
3 tasks
@whusym whusym closed this as completed Mar 6, 2018
@whusym whusym moved this from In Progress to Testing in Neuron Finding Mar 8, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
Neuron Finding
  
Testing
Development

No branches or pull requests

2 participants