diffstep 1.4.0

A symbolic differentiation library.

diffstep

diffstep is a small Python library for math derivatives.

It helps you:

• Find the derivative of math expressions.
• Get the derivative value at a given x value.
• Print a tree view of the expression.
• Print the steps used to get the derivative.

Installation

pip install diffstep

Quick Start

import diffstep

diffstep.diffstep("x^2 + 3x + 2")
# 2*x + 3

diffstep.gradient("x^2 + 3x + 2", 1)
# 5

diffstep.diffstep("x^x")
# x^x*(log(x) + 1)

diffstep.parse_expression("sin(x) + cos(x)")
# Prints pretty AST once and returns None

diffstep.trace("x^2 + sin(x^2)")
# Prints rendered trace text once and returns None

Public API

• diffstep(expression: str) Returns the simplified symbolic derivative as a SymPy expression.

• differentiate(expression: str) Alias of diffstep().

• gradient(expression: str, x_value: float) Differentiates, then evaluates at x = x_value.

• normalise_expression(expression: str) Validates and normalizes expression text (spaces, implicit multiplication, casing).

• tokenise_expression(expression: str) Returns the token list after normalization.

• parse_expression(expression: str, pretty: bool | None = None, emit: bool = True) If pretty=False, it returns the raw parse tree object. In other cases, it prints a readable tree and returns None. If you set emit=False, it does not print and returns the tree as text.

• pretty_parse_expression(expression: str, emit: bool = True) Convenience wrapper for parse_expression(expression, pretty=True).

• derived_ast(expression: str) Returns unsimplified derivative AST.

• trace(expression: str) Prints rendered differentiation trace text and returns None.

• DiffStep(expression: str) Deprecated alias of diffstep() (emits DeprecationWarning).

Supported Expressions

• Variables: x, e
• Numeric constants (integers and decimals)
• Operators: +, -, *, /, ^
• Parentheses and implicit multiplication (for example 2x, x(x+1))
• Power forms including:
  • x^n
  • a^f(x)
  • e^f(x)
  • General u(x)^v(x) (for example x^x)
• Functions:
  • sin, cos, tan
  • sec, cosec, cot
  • arcsin, arccos, arctan
  • ln, sqrt

Project Layout

• diffstep/core/: normalization, tokenization, parsing, AST nodes
• diffstep/differentiation/: derivative and trace rule engines
• diffstep/simplify/: SymPy-backed simplification
• diffstep/printer/: AST and trace rendering utilities
• diffstep/pipeline.py: user-facing orchestration API

License

MIT License. See LICENSE.txt.