scitex-gen 0.1.11

General utilities (caching, env detection, normalization, mat→npy, xml→dict, TimeStamper, …) — standalone module from the SciTeX ecosystem

SciTeX Gen (scitex-gen)

General-purpose Python utilities for scientific workflows

Full Documentation · uv pip install scitex-gen[all]

---

scitex-gen collects miscellaneous helpers that have been useful across many SciTeX projects: caching, normalization, environment detection, mat→npy conversion, xml→dict parsing, a TimeStamper, dimension wranglers, and more.

Problem and Solution

#	Problem	Solution
1	One-off utilities scatter across scripts -- every research repo grows its own to_even, xml2dict, TimeStamper, NaN-aware z-score	Curated misc bag -- scitex_gen.cache / to_z / to_nanz / xml2dict / TimeStamper / transpose / to_even / to_odd collected in one importable place
2	Environment detection ad-hoc -- "am I in a notebook? a script? interactive?" answered three different ways across projects	scitex_gen.detect_environment / is_notebook / is_script -- single canonical detector used across SciTeX peers
3	Legacy scitex.gen surface unstable -- helpers migrated to canonical peers (scitex_stats, scitex_os, scitex_str, ...) but old import paths still in use	Backward-compatible re-exports + DeprecationWarning -- scitex_gen.check_host etc. still work, point to the new home, and warn at import time

Installation

Requires Python >= 3.10.

pip install scitex-gen

Configuration

Copy .env.example to .env (or source it from your shell) to override defaults such as SCITEX_DIR.

Architecture

scitex_gen/
├── _fs/                  ← filesystem helpers (symlink, src, title2path, print_config)
├── _introspect/          ← introspection, caching, DimHandler, list_packages, mat2py, xml2dict
├── _ipython/             ← IPython/Jupyter helpers (notebook detection, embed, paste, less)
├── _legacy/              ← backward-compat wrappers (TimeStamper, deprecated start/close)
├── _numeric/             ← math/numerics (norm, norm_cache, symlog, to_even, to_odd, to_rank, transpose)
├── _skills/              ← agent-facing skill pages
├── misc.py               ← connect_nums, float_linspace
├── path.py               ← placeholder (path management lives in scitex-path)
├── _type.py / _var_info.py   ← type annotations (optional, requires xarray)
└── _wrap.py              ← text wrapping utility

Tiny single-purpose modules organized by domain. Most are pure Python; to_z and the geometric-median variants opt into torch when available. Several functions are re-exported from peer packages (scitex-context, scitex-os, scitex-sh, etc.) for backward compatibility.

1 Interfaces

Python API

import scitex_gen

scitex_gen.cache(...)
scitex_gen.TimeStamper()
scitex_gen.detect_environment()
scitex_gen.xml2dict(...)
scitex_gen.to_z(...)
scitex_gen.to_even(7)     # 6
scitex_gen.to_odd(8)      # 7
scitex_gen.transpose(...)
scitex_gen.list_packages()

Full API reference

Demo

flowchart LR
    A["raw tensor / array"] --> B["scitex_gen.to_z"]
    B --> C["NaN-aware z-scored output"]
    D["xml string"] --> E["scitex_gen.xml2dict"]
    E --> F["nested dict"]
    G["expensive func"] --> H["@scitex_gen.cache"]
    H --> I["disk-backed result"]

>>> import scitex_gen as gen
>>> gen.to_even(7)
6
>>> gen.to_odd(8)
7
>>> ts = gen.TimeStamper(); ts("step1")  # logs elapsed

Quick Start

import scitex_gen as gen

gen.cache(...)
gen.TimeStamper()
gen.xml2dict(...)
gen.to_z(tensor)          # requires torch
gen.to_even(n)
gen.to_odd(n)
gen.transpose(...)

Status

Standalone fork of scitex.gen. The umbrella package's scitex.gen import path is preserved via a sys.modules-alias bridge.

Decoupling notes:

• scitex.{decorators,str,os,introspect,session,context,sh,dict} → scitex_* direct imports (peer packages).
• scitex.torch.nanstd → optional via try/except with a torch-only fallback (only matters for _norm.to_z / to_nanz).
• import scitex removed from _less.py (was unused in module body).
• self-references in _norm_cache.py rewritten to scitex_gen.*.

Part of SciTeX

scitex-gen is part of SciTeX. Install via the umbrella with pip install scitex[gen] to use as scitex.gen (Python).

The SciTeX system follows the Four Freedoms for Research below, inspired by the Free Software Definition:

Four Freedoms for Research

0. The freedom to run your research anywhere — your machine, your terms.
1. The freedom to study how every step works — from raw data to final manuscript.
2. The freedom to redistribute your workflows, not just your papers.
3. The freedom to modify any module and share improvements with the community.

AGPL-3.0 — because we believe research infrastructure deserves the same freedoms as the software it runs on.

License

AGPL-3.0-only (see LICENSE).

---