Before you compute, print the layout. Shape mistakes are the usual cause of later ValueErrors.

Goal

Read shape, size, ndim, dtype, memory size, and a few reductions on a units grid.

Attributes

# Rows: Nairobi, Mombasa, Kisumu, Nakuru, Eldoret  ·  cols: A, B, C
units = np.array(
    [
        [12, 7, 2],
        [9, 6, 0],
        [3, 11, 1],
        [10, 8, 0],
        [6, 4, 0],
    ],
    dtype=float,
)
print("shape:", units.shape)
print("ndim:", units.ndim)
print("size:", units.size)
print("dtype:", units.dtype)
print("itemsize:", units.itemsize)
print("nbytes:", units.nbytes)

size is the product of the shape: 5 × 3 = 15.

Min, max, mean

units = np.array(
    [
        [12, 7, 2],
        [9, 6, 0],
        [3, 11, 1],
        [10, 8, 0],
        [6, 4, 0],
    ],
    dtype=float,
)
print("min:", np.min(units), "max:", np.max(units), "mean:", np.mean(units))
print("row 0 (Nairobi):", units[0])
print("col 1 (product B):", units[:, 1])

There is no pandas info() or describe() here. Print the pieces you care about.

Peek at the corners

units = np.array(
    [
        [12, 7, 2],
        [9, 6, 0],
        [3, 11, 1],
        [10, 8, 0],
        [6, 4, 0],
    ],
    dtype=float,
)
print("first two rows:")
print(units[:2])
print()
print("last row (Eldoret):", units[-1])

Tip

When a later chapter fails, print .shape of every array in the expression. Broadcasting errors name the shapes that could not stretch.