NeuroCarto package

Modules:

Probe Abstract:

neurocarto.probe

Probe Implementation:

neurocarto.probe_npx

UI components:

neurocarto.views

Utilities:

neurocarto.util

Symbols

The symbols and expressions used the document.

Numpy Array

Array[Dtype, *Shapes]

Examples

Array[int, 3, 2] is a 3x2 int number array.
Array[index:int, N] is a 1-d (N-length) int index array.
Array[int, N, (shank, col, row, state, category)] is a Nx5 int array with 5 columns (fields).
Array[float, [S,], R, C] is either a 2-d (shape (R, C)) or a 3-d (shape (S, R, C)) float array.
Array[V, ..., N, ...] is a multi-dimension V-domain array, which has one N-length axis.
Array[bool, N] is a bool N-length array, often named as a mark.

Use examples

An array is applied with a mask:

a = np.arange(10)
a = a[a > 5]

We can document the __getitem__ of a numpy array as:

def __getitem__(self: Array[V, N], mark: Array[bool, N]) -> Array[V, M]:
    """
    V : any value type
    N : N-length array
    M = np.count_nonzero(mask)
    """

Some transform

a: Array[int, *A]: a.shape: A, a.shape[i] == A[i]
a: Array[int, A, B]: a.T: Array[int, B, A]
a: Array[int, N, (A, B, C)]: a[:, 1]: Array[B:int, N], a[:, [1]]: Array[B:int, N, 1]
a: Array[int, A] and b: Array[A, *B]: a[b]: Array[int, *B]
a: Array[int, A] and b: Array[bool, A]: a[b]: Array[int, A']

where A' = np.count_nonzero(b)
a: Array[int, A, B]: np.sum(a, axis=0) : Array[int, B]