Decimal Mathematics

[rhubarbdog]

if floating point numbers are letting you down because of the gaps and rounding errors try my decimal numbers class decimal.py at https://github.com/rhubarbdog/decimal-numbers . There is a version for microbit called microbit.py follow the instructions in the README.md.
See example code decimal_math.py and pi.py for python 3 and other boards and for microbit it's microbit_pi.py microbit_root.py

[sosi-deadeye]

I think you should also post your module here: https://github.com/mcauser/awesome-micropython

[jimmo]

@rhubarbdog This is very cool!

Just a question about the implementation though -- you've essentially implemented arbitrary precision integers from scratch in Python (which is neat!!). But MicroPython already does this for you -- by default on almost all ports, integers have no limit (i.e. MICROPY_LONGINT_IMPL == MICROPY_LONGINT_IMPL_MPZ).

So could a decimal number (i.e. your Number class) just be a python integer and a scale (i.e. offset of the decimal point).

Also, did you consider implementing the arithmetic operators __add__ etc on Number, rather than having an external decimal.add() function?

[dhalbert]

Another related library for CircuitPython, by @jepler: https://github.com/jepler/Jepler_CircuitPython_udecimal

[peterhinch]

See also https://github.com/mpy-dev/micropython-decimal-number