structured-array 0.0.1

Efficient manipulation of the numpy structured arrays.

structured-array

Efficient manipulation of the numpy structured arrays.

---

Installation

pip install structured-array

Examples

A structured array can easily be created.

import structured_array as st

arr = st.array({
    "label": ["a", "b", "c"],
    "value": [4, 5, 6],
    "array": [np.zeros(3), np.ones(3), np.zeros(3)],
})
arr

label    value    array
[<U1]    [<i8]    [<f8]
-------  -------  ----------
a        4        (3,) array
b        5        (3,) array
c        6        (3,) array

You can directly read and write the structured array from/to a npy file.

arr = st.read_npy("data.npy")

structured-array use polars expression system to manipulate the structured array.

arr.select("label", st.col("value") + 1)  # column selection
arr.group_by("label").agg(st.col("value").sum())  # aggregation
arr.filter(st.col("value") > 5)  # filtering

License

structured-array is distributed under the terms of the MIT license.