Snuggs

1.4.7 · active · verified Mon Apr 13

Snuggs is a Python library that provides s-expressions for Numpy, allowing users to define and evaluate array computations using a Lisp-like syntax. It is currently at version 1.4.7 and appears to have a stable, though not frequently updated, release cadence, with the last release in September 2019.

Warnings

deprecated The `kwd_dict` parameter in `snuggs.eval()` is deprecated. Users should pass context variables directly as keyword arguments instead.
Fix: Replace `snuggs.eval(expression, kwd_dict={'key': value})` with `snuggs.eval(expression, key=value)`.
gotcha Snuggs does not provide performance optimizations like multithreading or elimination of temporary data, unlike libraries such as `numexpr`. It is primarily designed for simple calculator programs.
Fix: For performance-critical array computations, consider using libraries explicitly designed for optimized numerical operations or vectorized NumPy code directly, rather than relying on Snuggs for performance gains.
gotcha When using functions or operators that reference the order in which values have been provided within the evaluation context (e.g., 'read' function), it's important to pass a dictionary, specifically an `OrderedDict`, as keyword arguments might not preserve order.
Fix: For order-dependent context, create an `OrderedDict` and pass its items as keyword arguments: `ctx = OrderedDict((('a', np.array([5, 5])), ('b', np.array([2, 2])))); snuggs.eval('(- (read 1) (read 2))', **ctx)`.

Install

pip install snuggs Install stable version

Imports

snuggs
```
import snuggs
```
The primary way to interact with the library is through the top-level 'snuggs' module.

Quickstart

This quickstart demonstrates basic arithmetic operations, array creation using `asarray`, and evaluating expressions with a local context. The `snuggs.eval()` function is central to executing s-expressions.

import snuggs
import numpy as np

# Basic arithmetic
result_add = snuggs.eval('(+ 1 2)')
print(f"Addition: {result_add}")

# Array creation and multiplication
result_multiply_array = snuggs.eval("(* 3.5 (asarray 1 1))")
print(f"Multiply with array: {result_multiply_array}")

# Evaluation with a local context
ctx_array = np.array([2, 2])
result_context = snuggs.eval("(+ (asarray 1 1) b)", b=ctx_array)
print(f"With context: {result_context}")

view raw JSON →