tutorial nb

.github/workflows/docker-image.yml

@@ -1,26 +1,26 @@
-name: Docker Image CI
+# name: Docker Image CI
 
-on:
-  push:
-    branches: [ "develop" ]
-  pull_request:
-    branches: [ "develop" ]
+# on:
+#   push:
+#     branches: [ "develop" ]
+#   pull_request:
+#     branches: [ "develop" ]
 
-jobs:
+# jobs:
 
-  build:
+#   build:
 
-    runs-on: ubuntu-latest
+#     runs-on: ubuntu-latest
 
-    steps:
-    - uses: actions/checkout@v3
-    - name: docker login
-      env:
-        DOCKER_USER: ${{secrets.DOCKER_USER}}
-        DOCKER_PASSWORD: ${{secrets.DOCKER_PASSWORD}}
-      run: |
-        docker login -u $DOCKER_USER -p $DOCKER_PASSWORD
-    - name: Build the Docker image
-      run: docker build . --file Dockerfile --tag neurodata/brainlit:latest
-    - name: Docker Push
-      run: docker push neurodata/brainlit
+#     steps:
+#     - uses: actions/checkout@v3
+#     - name: docker login
+#       env:
+#         DOCKER_USER: ${{secrets.DOCKER_USER}}
+#         DOCKER_PASSWORD: ${{secrets.DOCKER_PASSWORD}}
+#       run: |
+#         docker login -u $DOCKER_USER -p $DOCKER_PASSWORD
+#     - name: Build the Docker image
+#       run: docker build . --file Dockerfile --tag neurodata/brainlit:latest
+#     - name: Docker Push
+#       run: docker push neurodata/brainlit

brainlit/map_neurons/diffeo_gen.py

@@ -151,14 +151,14 @@ def diffeo_gen_ara(sigma):
     """Return random diffeomorphism generated by sampling Gaussian noise then passing through Riemannian exponential.
 
     Args:
-        sigma (float): standard deviation of noise.
+        sigma (float): standard deviation of noise in microns.
 
     Returns:
         List: list of sampled points in spatial domain.
         np.array: range of diffeomorphism at sampled points.
     """
     # a domain for sampling your velocity and deformatoin
-    dv = np.array([5.0, 5.0, 5.0])
+    dv = np.array([100.0, 100.0, 100.0])  # units are every 100 microns
     nv = np.array([132, 80, 114])
     xv = [np.arange(n) * d - (n - 1) * d / 2 for n, d in zip(nv, dv)]
 
@@ -169,7 +169,7 @@ def diffeo_gen_ara(sigma):
     # a frequency domain
     fv = [np.arange(n) / n / d for n, d in zip(nv, dv)]
     FV = np.stack(np.meshgrid(*fv, indexing="ij"), -1)
-    a = 5.0
+    a = 100.0
     p = 2.0
     LL = (
         1.0 - 2.0 * a**2 * np.sum(((np.cos(2.0 * np.pi * FV * dv) - 1)) / dv**2, -1)
@@ -189,4 +189,4 @@ def diffeo_gen_ara(sigma):
     phii = expR([torch.tensor(x) for x in xv], torch.tensor(v), K, n=10)
     phii = phii.detach().cpu().numpy()
 
-    return xv, phii
+    return xv, phii, K

docs/notebooks/pipelines/BrainLine/validation-data/soma/brainr_results/image_2000_1900_1800_somas.txt

@@ -0,0 +1,3 @@
+[2008.3863216266175, 1994.6025878003697, 1840.358595194085]
+[2062.8153241650293, 1949.3064833005894, 1822.9626719056973]
+[2068.915071770335, 1930.4234449760766, 1856.846889952153]

experiments/map_neurons/README.rst

@@ -10,9 +10,9 @@ Neuron Mapping
 
 * The notebook that best demonstrates the code in this project is: ``brainlit/experiments/map_neurons/tutorial.ipynb``.
 
-* As a general rule, figures were made in the notebooks in ``brainlit/experiments/map_neurons/`` and, when necessary, heavy computations were done using scripts in ``brainlit/experiments/map_neurons/other``. Many of the notebooks rely on data generated by scripts and therefore will not run as-is.
+* Figures in tha paper were made in the notebooks in ``brainlit/experiments/map_neurons/`` and, when necessary, heavy computations were done using scripts in ``brainlit/experiments/map_neurons/other``. Many of the notebooks rely on data generated by scripts and therefore will not run as-is.
 
-* The notebooks that made the specific figures in the current version of the manuscript are:
+* Nonetheless, the notebooks can serve as a reference to the interested reader. The notebooks that made the specific figures in the current version of the manuscript are:
 
 * ``map_neurons.ipynb``: Fig. 1a-b, 2.