Write arbitrary NumPy scripts, simulate broadcasting stretches, watch axis collapse arrows, or process image tensors.
Create a 1D ndarray vector from a standard Python list.
Create a 2D matrix (rows x cols) from nested Python lists.
Pre-allocate memory with zeros in specified (rows, cols) shape.
Pre-allocate memory filled with float 1.0 values.
Create an array pre-filled with a specific constant value (e.g. 7).
Create an identity matrix with 1s on main diagonal and 0s elsewhere.
Construct a 2D diagonal matrix from 1D numbers or extract diagonal.
Generate a sequence from start to stop with custom step jump.
Generate N perfectly spaced fractional samples between bounds.
Generate samples spaced evenly on a log scale (10^start to 10^stop).
Algebraic dot-product matrix multiplication (A @ B).
Compute inner product of vectors or algebraic matrix product.
Calculate matrix determinant |A| = (ad - bc) for scaling factor.
Compute multiplicative inverse such that A @ A^-1 = Identity.
Flip matrix over its main diagonal, switching rows and columns.
Calculate sum of elements along the primary matrix diagonal.
Calculate Euclidean vector length sqrt(x1^2 + x2^2 + ...).
Compute the outer product matrix of two vectors u and v.
Calculate perpendicular cross product vector in 3D Euclidean space.
See how NumPy broadcasts a single scalar value across all elements.
Broadcast a 1D vector across rows of a 2D matrix.
Extract a sub-array window excluding stop index.
Reverse array order using negative step slicing.
Extract a 2x2 submatrix block from a 3x3 grid.
Sample uniform floating point numbers in [0.0, 1.0) with custom dimensions.
Standard Normal Distribution (Mean=0, Variance=1) with negative & positive floats.
Sample random discrete integers in [low, high) with specified shape size.
Sample continuous floats in [0.0, 1.0) by passing a shape tuple.
Sample random elements from a discrete array or list of choices.
Randomly shuffle the elements of an existing array in-place.
Return a newly shuffled copy or permute a sequence from 0 to N-1.
Sample continuous floats uniformly distributed between custom boundaries.
Sample values from a Gaussian distribution with custom loc (mean) and scale (std).
Collapses row dimension to calculate column totals.
Collapses column dimension to calculate row totals.
Apply vectorized square root and exponential functions.
Detect and replace negative values with 0 using np.where.