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.

Sign up for GitHub

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

JR SC variable rename #159

Draft
wants to merge 10 commits into
base: dev
Choose a base branch
from
Draft

JR SC variable rename #159

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
160c81d
Update docs.yml
JohnWangDataAnalyst Feb 21, 2023
d6dfef0
Merge pull request #93 from GriffithsLab/dev
JohnWangDataAnalyst Jul 25, 2023
d32b507
Fix: Change docs build ref to `main`
tmorshed Sep 6, 2023
124454e
Merge pull request #121 from GriffithsLab/tmorshed/enh/sphinx-to-main
Andrew-Clappison Sep 11, 2023
a533053
Merge pull request #122 from GriffithsLab/dev
tmorshed Sep 14, 2023
ef921c3
Merge pull request #137 from GriffithsLab/dev
Davi1990 Nov 21, 2023
2798469
Update jansen_rit.py
Minarose Jan 17, 2024
e7c438b
Update jansen_rit.py gains
Minarose Jan 17, 2024
6d9b0f3
rename sc_fitted in eg004r__fitting_JR_example.py
Minarose Jan 17, 2024
7e011a4
Merge branch 'GriffithsLab:main' into JR_sc
Minarose Jan 17, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions examples/eg004r__fitting_JR_example.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -154,7 +154,7 @@
im0 = ax[0].imshow(sc, cmap='bwr', vmin = 0.0, vmax = 0.02)
ax[0].set_title('The empirical SC')
fig.colorbar(im0, ax=ax[0], fraction=0.046, pad=0.04)
im1 = ax[1].imshow(F.model.sc_fitted.detach().numpy(), cmap='bwr', vmin = 0.0, vmax = 0.02)
im1 = ax[1].imshow(F.model.sc_pp.detach().numpy(), cmap='bwr', vmin = 0.0, vmax = 0.02)
ax[1].set_title('The fitted SC')
fig.colorbar(im1, ax=ax[1], fraction=0.046, pad=0.04)
plt.show()
Expand All @@ -168,4 +168,4 @@
ax[1].set_title('Test')
ax[2].plot(eeg_data.T)
ax[2].set_title('empirical')
plt.show()
plt.show()
68 changes: 34 additions & 34 deletions whobpyt/models/JansenRit/jansen_rit.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -203,21 +203,21 @@ def setModelParameters(self):
param_reg = []
param_hyper = []

# Set w_bb, w_ff, and w_ll as attributes as type Parameter if use_fit_gains is True
# Set w_pi, w_pe, and w_pp as attributes as type Parameter if use_fit_gains is True
if self.use_fit_gains:
self.w_bb = Parameter(torch.tensor(np.zeros((self.node_size, self.node_size)) + 0.05, # the backwards gains
self.w_pi = Parameter(torch.tensor(np.zeros((self.node_size, self.node_size)) + 0.05, # the backwards gains
dtype=torch.float32))
self.w_ff = Parameter(torch.tensor(np.zeros((self.node_size, self.node_size)) + 0.05, # the forward gains
self.w_pe = Parameter(torch.tensor(np.zeros((self.node_size, self.node_size)) + 0.05, # the forward gains
dtype=torch.float32))
self.w_ll = Parameter(torch.tensor(np.zeros((self.node_size, self.node_size)) + 0.05, # the lateral gains
self.w_pp = Parameter(torch.tensor(np.zeros((self.node_size, self.node_size)) + 0.05, # the lateral gains
dtype=torch.float32))
param_reg.append(self.w_ll)
param_reg.append(self.w_ff)
param_reg.append(self.w_bb)
param_reg.append(self.w_pp)
param_reg.append(self.w_pe)
param_reg.append(self.w_pi)
else:
self.w_bb = torch.tensor(np.zeros((self.node_size, self.node_size)), dtype=torch.float32)
self.w_ff = torch.tensor(np.zeros((self.node_size, self.node_size)), dtype=torch.float32)
self.w_ll = torch.tensor(np.zeros((self.node_size, self.node_size)), dtype=torch.float32)
self.w_pi = torch.tensor(np.zeros((self.node_size, self.node_size)), dtype=torch.float32)
self.w_pe = torch.tensor(np.zeros((self.node_size, self.node_size)), dtype=torch.float32)
self.w_pp = torch.tensor(np.zeros((self.node_size, self.node_size)), dtype=torch.float32)

# If use_fit_lfm is True, set lm as an attribute as type Parameter (containing variance information)
if self.use_fit_lfm:
Expand Down Expand Up @@ -329,32 +329,32 @@ def forward(self, external, hx, hE):

if self.sc.shape[0] > 1:

# Update the Laplacian based on the updated connection gains w_bb.
w_b = torch.exp(self.w_bb) * torch.tensor(self.sc, dtype=torch.float32)
w_n_b = w_b / torch.linalg.norm(w_b)
self.sc_m_b = w_n_b
dg_b = -torch.diag(torch.sum(w_n_b, dim=1))

# Update the Laplacian based on the updated connection gains w_ff.
w_f = torch.exp(self.w_ff) * torch.tensor(self.sc, dtype=torch.float32)
w_n_f = w_f / torch.linalg.norm(w_f)
self.sc_m_f = w_n_f
dg_f = -torch.diag(torch.sum(w_n_f, dim=1))

# Update the Laplacian based on the updated connection gains w_ll.
w_l = torch.exp(self.w_ll) * torch.tensor(self.sc, dtype=torch.float32)
w_n_l = (0.5 * (w_l + torch.transpose(w_l, 0, 1))) / torch.linalg.norm(
0.5 * (w_l + torch.transpose(w_l, 0, 1)))
self.sc_fitted = w_n_l
dg_l = -torch.diag(torch.sum(w_n_l, dim=1))
# Update the Laplacian based on the updated connection gains w_pi.
w_i= torch.exp(self.w_pi) * torch.tensor(self.sc, dtype=torch.float32)
w_n_pi = w_i/ torch.linalg.norm(w_i)
self.sc_p_i = w_n_pi
dg_b = -torch.diag(torch.sum(w_n_pi, dim=1))

# Update the Laplacian based on the updated connection gains w_pe.
w_e = torch.exp(self.w_pe) * torch.tensor(self.sc, dtype=torch.float32)
w_n_pe = w_e / torch.linalg.norm(w_e)
self.sc_p_e = w_n_pe
dg_f = -torch.diag(torch.sum(w_n_pe, dim=1))

# Update the Laplacian based on the updated connection gains w_pp.
w_p = torch.exp(self.w_pp) * torch.tensor(self.sc, dtype=torch.float32)
w_n_pp = (0.5 * (w_p + torch.transpose(w_p, 0, 1))) / torch.linalg.norm(
0.5 * (w_p + torch.transpose(w_p, 0, 1)))
self.sc_p_p = w_n_pp
dg_l = -torch.diag(torch.sum(w_n_pp, dim=1))
else:
l_s = torch.tensor(np.zeros((1, 1)), dtype=torch.float32) #TODO: This is not being called anywhere
dg_l = 0
dg_b = 0
dg_f = 0
w_n_l = 0
w_n_b = 0
w_n_f = 0
w_n_pp = 0
w_n_pi = 0
w_n_pe = 0

self.delays = (self.dist / mu).type(torch.int64)

Expand All @@ -376,11 +376,11 @@ def forward(self, external, hx, hE):
Ed = torch.tensor(np.zeros((self.node_size, self.node_size)), dtype=torch.float32) # delayed E
hE_new = hE.clone()
Ed = hE_new.gather(1, self.delays)
LEd_b = torch.reshape(torch.sum(w_n_b * torch.transpose(Ed, 0, 1), 1),
LEd_b = torch.reshape(torch.sum(w_n_pi * torch.transpose(Ed, 0, 1), 1),
(self.node_size, 1))
LEd_f = torch.reshape(torch.sum(w_n_f * torch.transpose(Ed, 0, 1), 1),
LEd_f = torch.reshape(torch.sum(w_n_pe * torch.transpose(Ed, 0, 1), 1),
(self.node_size, 1))
LEd_l = torch.reshape(torch.sum(w_n_l * torch.transpose(Ed, 0, 1), 1),
LEd_l = torch.reshape(torch.sum(w_n_pp * torch.transpose(Ed, 0, 1), 1),
(self.node_size, 1))

# TMS (or external) input
Expand Down
Loading