MicroPython
MicroPython is a small implementation of a subset of Python 3 intended for use in microcontrollers. MicroPython can also be embedded as a scripting engine in other applications. See the MicroPython Website.
Packages
femm-qt's implementation of MicroPython supports the following packages:
1. Compiled into femm-qt (22)
No files, always present:
- femm
- numpy
- scipy
- ulab
- math
- cmath
- array
- struct
- collections
- io
- os
- sys
- gc
- time
- json
- random
- binascii
- hashlib
- errno
- builtins
- micropython
numpy and scipy here are ulab, not the real packages — a deliberate subset. In-process numpy exposes 122 names including linalg, fft and random; scipy gives integrate, linalg, optimize, signal, special. Enough for ordinary array work and the linear algebra femm-qt added on top (complex eig/qr/svd/solve, an int32 dtype), but not the full API.
2. Pure Python, shipped and on sys.path (30)
- abc
- argparse
- base64
- bisect
- contextlib
- copy
- csv
- datetime
- fnmatch
- functools
- glob
- heapq
- inspect
- itertools
- logging
- operator
- pathlib
- pickle
- pprint
- re
- shutil
- stat
- string
- tempfile
- textwrap
- traceback
- types
- ucontextlib
sys.path is '' | <app>/bin/python | %APPDATA%\FEMM\python — that last one is yours, for your own modules.
3. Bridged to CPython (3+)
The packages:
- matplotlib
- xlsxwriter
- zipfile
- pandas
are bridged to the local CPython install, but they are shimmed in such a way that they appeared to be installed locally. For example,
import matplotlib works.
Other packages that do not have shims are still available through pybridge.import_module(), which takes any module name. For example:
import pybridge
np = pybridge.import_module(“numpy”) # the real one, not ulab
opt = pybridge.import_module(“scipy.optimize”)