More updates

README.rst

@@ -1,6 +1,5 @@
 
 
-
 .. image:: https://github.com/PKU-NIP-Lab/BrainPy/blob/master/docs/images/logo.png
     :target: https://github.com/PKU-NIP-Lab/BrainPy
     :align: center
@@ -65,8 +64,11 @@ Install ``BrainPy`` using ``pip``::
 
 Install from source code::
 
+    > git clone https://github.com/PKU-NIP-Lab/BrainPy
     > python setup.py install
+    >
     > # or
+    >
     > pip install git+https://github.com/PKU-NIP-Lab/BrainPy
 
 

brainpy/connectivity/formatter.py

@@ -2,13 +2,22 @@
 
 from .. import numpy as np
 from .. import profile
+from ..errors import ModelUseError
 
 try:
     import numba as nb
 except ImportError:
     nb = None
 
-__all__ = ['mat2ij', 'pre2post', 'pre2syn', 'post2syn', 'post2pre']
+__all__ = [
+    'mat2ij',
+    'pre2post',
+    'pre2syn',
+    'post2syn',
+    'post2pre',
+    'post_slicing_by_pre',
+    'pre_slicing_by_post',
+]
 
 
 def mat2ij(conn_mat):
@@ -17,23 +26,20 @@ def mat2ij(conn_mat):
     Parameters
     ----------
     conn_mat : numpy.ndarray
-        Connectivity matrix with `(num_pre x num_post)` shape.
+        Connectivity matrix with `(num_pre, num_post)` shape.
 
     Returns
     -------
     conn_tuple : tuple
         (Pre-synaptic neuron indexes,
          post-synaptic neuron indexes).
     """
-    pre_ids = []
-    post_ids = []
-    num_pre = conn_mat.shape[0]
-    for pre_idx in range(num_pre):
-        post_idxs = np.where(conn_mat[pre_idx] > 0)[0]
-        post_ids.extend(post_idxs)
-        pre_ids.extend([pre_idx] * len(post_idxs))
-    pre_ids = np.array(pre_ids, dtype=np.uint64)
-    post_ids = np.array(post_ids, dtype=np.uint64)
+    conn_mat = np.asarray(conn_mat)
+    try:
+        assert np.ndim(conn_mat) == 2
+    except AssertionError:
+        raise ModelUseError('Connectivity matrix must be in the shape of (num_pre, num_post).')
+    pre_ids, post_ids = np.where(conn_mat > 0)
     return pre_ids, post_ids
 
 
@@ -42,34 +48,32 @@ def pre2post(i, j, num_pre):
 
     Parameters
     ----------
-    i : a_list, numpy.ndarray
+    i : list, numpy.ndarray
         The pre-synaptic neuron indexes.
-    j : a_list, numpy.ndarray
+    j : list, numpy.ndarray
         The post-synaptic neuron indexes.
     num_pre : int, None
         The number of the pre-synaptic neurons.
 
     Returns
     -------
     conn : list
-        The conn a_list of pre2post.
+        The conn list of pre2post.
     """
-    i, j = np.asarray(i), np.asarray(j)
+    try:
+        assert len(i) == len(j)
+    except AssertionError:
+        raise ModelUseError('The length of "i" and "j" must be the same.')
+
+    pre2post_list = [[] for _ in range(num_pre)]
+    for pre_id, post_id in zip(i, j):
+        pre2post_list[pre_id].append(post_id)
 
     if profile.is_numba_bk():
-        pre2post_list = nb.typed.List()
-
-        for pre_i in range(num_pre):
-            index = np.where(i == pre_i)[0]
-            post_idx = j[index]
-            pre2post_list.append(np.uint64(post_idx))
-    else:
-        pre2post_list = []
-
-        for pre_i in range(num_pre):
-            index = np.where(i == pre_i)[0]
-            post_idx = j[index]
-            pre2post_list.append(np.uint64(post_idx))
+        pre2post_list_nb = nb.typed.List()
+        for pre_id in range(num_pre):
+            pre2post_list_nb.append(np.int64(pre2post_list[pre_id]))
+        pre2post_list = pre2post_list_nb
     return pre2post_list
 
 
@@ -78,94 +82,167 @@ def post2pre(i, j, num_post):
 
     Parameters
     ----------
-    i : a_list, numpy.ndarray
+    i : list, numpy.ndarray
         The pre-synaptic neuron indexes.
-    j : a_list, numpy.ndarray
+    j : list, numpy.ndarray
         The post-synaptic neuron indexes.
     num_post : int, None
         The number of the post-synaptic neurons.
 
     Returns
     -------
     conn : list
-        The conn a_list of post2pre.
+        The conn list of post2pre.
     """
-    i, j = np.asarray(i), np.asarray(j)
+
+    try:
+        assert len(i) == len(j)
+    except AssertionError:
+        raise ModelUseError('The length of "i" and "j" must be the same.')
+
+    post2pre_list = [[] for _ in range(num_post)]
+    for pre_id, post_id in zip(i, j):
+        post2pre_list[post_id].append(pre_id)
 
     if profile.is_numba_bk():
-        post2pre_list = nb.typed.List()
-        for post_i in range(num_post):
-            index = np.where(j == post_i)[0]
-            pre_idx = i[index]
-            post2pre_list.append(np.uint64(pre_idx))
-    else:
-        post2pre_list = []
-        for post_i in range(num_post):
-            index = np.where(j == post_i)[0]
-            pre_idx = i[index]
-            post2pre_list.append(np.uint64(pre_idx))
+        post2pre_list_nb = nb.typed.List()
+        for post_id in range(num_post):
+            post2pre_list_nb.append(np.int64(post2pre_list[post_id]))
+        post2pre_list = post2pre_list_nb
     return post2pre_list
 
 
-def pre2syn(i, j, num_pre):
+def pre2syn(i, num_pre):
     """Get pre2syn connections from `i` and `j` indexes.
 
     Parameters
     ----------
-    i : a_list, numpy.ndarray
+    i : list, numpy.ndarray
         The pre-synaptic neuron indexes.
-    j : a_list, numpy.ndarray
-        The post-synaptic neuron indexes.
     num_pre : int
         The number of the pre-synaptic neurons.
 
     Returns
     -------
     conn : list
-        The conn a_list of pre2syn.
+        The conn list of pre2syn.
     """
-    i, j = np.asarray(i), np.asarray(j)
+    pre2syn_list = [[] for _ in range(num_pre)]
+    for syn_id, pre_id in enumerate(i):
+        pre2syn_list[pre_id].append(syn_id)
 
     if profile.is_numba_bk():
-        post2syn_list = nb.typed.List()
-        for pre_i in range(num_pre):
-            index = np.where(i == pre_i)[0]
-            post2syn_list.append(np.uint64(index))
-    else:
-        post2syn_list = []
-        for pre_i in range(num_pre):
-            index = np.where(i == pre_i)[0]
-            post2syn_list.append(np.uint64(index))
-    return post2syn_list
+        pre2syn_list_nb = nb.typed.List()
+        for pre_ids in pre2syn_list:
+            pre2syn_list_nb.append(np.int64(pre_ids))
+        pre2syn_list = pre2syn_list_nb
+    return pre2syn_list
 
 
-def post2syn(i, j, num_post):
+def post2syn(j, num_post):
     """Get post2syn connections from `i` and `j` indexes.
 
     Parameters
     ----------
-    i : a_list, numpy.ndarray
-        The pre-synaptic neuron indexes.
-    j : a_list, numpy.ndarray
+    j : list, numpy.ndarray
         The post-synaptic neuron indexes.
     num_post : int
         The number of the post-synaptic neurons.
 
     Returns
     -------
     conn : list
-        The conn a_list of post2syn.
+        The conn list of post2syn.
     """
-    i, j = np.asarray(i), np.asarray(j)
+    post2syn_list = [[] for _ in range(num_post)]
+    for syn_id, post_id in enumerate(j):
+        post2syn_list[post_id].append(syn_id)
 
     if profile.is_numba_bk():
-        post2syn_list = nb.typed.List()
-        for post_i in range(num_post):
-            index = np.where(j == post_i)[0]
-            post2syn_list.append(np.uint64(index))
-    else:
-        post2syn_list = []
-        for post_i in range(num_post):
-            index = np.where(j == post_i)[0]
-            post2syn_list.append(np.uint64(index))
+        post2syn_list_nb = nb.typed.List()
+        for pre_ids in post2syn_list:
+            post2syn_list_nb.append(np.int64(pre_ids))
+        post2syn_list = post2syn_list_nb
     return post2syn_list
+
+
+def post_slicing_by_pre(i, j, num_pre):
+    """Get post slicing connections by pre-synaptic ids.
+
+    Parameters
+    ----------
+    i : list, numpy.ndarray
+        The pre-synaptic neuron indexes.
+    j : list, numpy.ndarray
+        The post-synaptic neuron indexes.
+    num_pre : int
+        The number of the pre-synaptic neurons.
+
+    Returns
+    -------
+    conn : list
+        The conn list of post2syn.
+    """
+    try:
+        assert len(i) == len(j)
+    except AssertionError:
+        raise ModelUseError('The length of "i" and "j" must be the same.')
+
+    # pre2post connection
+    pre2post_list = [[] for _ in range(num_pre)]
+    for pre_id, post_id in zip(i, j):
+        pre2post_list[pre_id].append(post_id)
+    post_indices = np.concatenate(pre2post_list)
+    post_indices = np.asarray(post_indices, dtype=np.int_)
+
+    # pre2post slicing
+    post_slicing = []
+    start = 0
+    for post_ids in pre2post_list:
+        end = start + len(post_ids)
+        post_slicing.append([start, end])
+        start = end
+    post_slicing = np.asarray(post_slicing, dtype=np.int_)
+
+    return post_indices, post_slicing
+
+
+def pre_slicing_by_post(i, j, num_post):
+    """Get pre slicing connections by post-synaptic ids.
+
+    Parameters
+    ----------
+    i : list, numpy.ndarray
+        The pre-synaptic neuron indexes.
+    j : list, numpy.ndarray
+        The post-synaptic neuron indexes.
+    num_post : int
+        The number of the post-synaptic neurons.
+
+    Returns
+    -------
+    conn : list
+        The conn list of post2syn.
+    """
+    try:
+        assert len(i) == len(j)
+    except AssertionError:
+        raise ModelUseError('The length of "i" and "j" must be the same.')
+
+    # post2pre connection
+    post2pre_list = [[] for _ in range(num_post)]
+    for pre_id, post_id in zip(i, j):
+        post2pre_list[post_id].append(pre_id)
+    pre_indices = np.concatenate(post2pre_list)
+    pre_indices = np.asarray(pre_indices, dtype=np.int_)
+
+    # post2pre slicing
+    pre_slicing = []
+    start = 0
+    for pre_ids in post2pre_list:
+        end = start + len(pre_ids)
+        pre_slicing.append([start, end])
+        start = end
+    pre_slicing = np.asarray(pre_slicing, dtype=np.int_)
+
+    return pre_indices, pre_slicing