Ready to use with all neuron types for CNN

[brando90]

has all neurons types:

• filter so each filter is one neuron and data patch is a data point. neuron vector of size [BHW, 1] with C of them.
• activation so each activation is one neuron and the neuron vector is of size [B, 1] with [CHW] of them
• original anatome too, but don't have a full understanding of why it's shaped the way it is but you can use it.

Sanity checks for all work. See jupyter notebooks.

---

This discussion was created from the release Ready to use with all neuron types for CNN.

[brando90]

doesn't seem like the interpretation of activation of filter made a huge difference...curious! Interesting. Further experiments some day to confirm this could be nice.

Seems all results are the same...ok! Some have tiny data so unreliable but others do not and are indistinguishable. Good so far.

---

Should be very very close to 1.0: sim=1.0 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.0 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.000000238418579 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=0.9999998807907104 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=0.9999998807907104 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.000000238418579 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.0 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.0000000381469718 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.000000238418579 (cxa_dist_type='pwcca')
import sys; print('Python %s on %s' % (sys.version, sys.platform))
Python 3.9.6 (default, Aug 18 2021, 12:38:10) 
Type 'copyright', 'credits' or 'license' for more information
IPython 7.28.0 -- An enhanced Interactive Python. Type '?' for help.
PyDev console: using IPython 7.28.0
Python 3.9.6 (default, Aug 18 2021, 12:38:10) 
[Clang 10.0.0 ] on darwin

[brando90]

Should be very very close to 1.0: sim=1.0 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.0 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.000000238418579 (cxa_dist_type='pwcca')
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 32768]), torch.Size([4, 32768])
Should be very very close to 1.0: sim=0.9999998807907104 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=0.9999998807907104 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.000000238418579 (cxa_dist_type='pwcca')
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 200]), torch.Size([4, 200])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([8, 4096]), torch.Size([8, 4096])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([8, 4096]), torch.Size([8, 4096])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([8, 4096]), torch.Size([8, 4096])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([8, 4096]), torch.Size([8, 4096])
Should be very very close to 1.0: sim=1.0 (cxa_dist_type='pwcca')
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
WARNING:root:Warning, you have less data points than features in one of your data matrices: torch.Size([4, 8]), torch.Size([4, 8])
Should be very very close to 1.0: sim=1.0000000381469718 (cxa_dist_type='pwcca')
Should be very very close to 1.0: sim=1.000000238418579 (cxa_dist_type='pwcca')
import sys; print('Python %s on %s' % (sys.version, sys.platform))
Python 3.9.6 (default, Aug 18 2021, 12:38:10) 
Type 'copyright', 'credits' or 'license' for more information
IPython 7.28.0 -- An enhanced Interactive Python. Type '?' for help.
PyDev console: using IPython 7.28.0
Python 3.9.6 (default, Aug 18 2021, 12:38:10) 
[Clang 10.0.0 ] on darwin