NumpSy

Straight up mix between NumPy, SymPy and Pandas into a value single-declaration extendable framework to simulatenously perform symbolic and numerical operations.

Objectives:

1. Ever think you wanted to simultaneously perform numerical and symbolic mathematics for an engineering or optimization derivation? Now you can pretty much intuitively derivate simultaneously whilst performing unit management automatically.
2. Integrate mathematical analytical derivation Python toolchains into a single handy one that retains and expands each of the constituent packages methods. Retain intuitive compatibility.
3. Have fun!

Is it any good?

I think it's an elegant mathematical representation to simultanously perform symbolic, numerical, and data science operations into a single system. It targets a minimal overhead to raw numpy, sympy or scipy operations.

Quick Start

Download the Anaconda distribution first.

Pip install:

$ pip install numpsy

Local install for most recent version:

$ git clone https://github.com/daquintero/numpsy.git
$ cd numpsy
$ python3 setup.py install

Quick Example

See the 10 minutes to NumpSy jupyter notebook for much more.

Installation

Import NumpSy

import numpsy as nsy

Units

Declare a Unit

meter = nsy.Unit(name="meter", symbol="m")
meter

Retrieve attributes from this Unit

meter.s

$\displaystyle m$

meter.symbol

$\displaystyle m$

meter.name

'meter'

Operate with this unit

farad_per_meter = nsy.Unit(name="Farad", symbol="F") / meter
farad_per_meter

Append to Unit Library

nsy.Units().data

Hertz     Unit       name name_expression               ...
Farad     Unit       name name_expression               ...
meter     Unit       name name_expression               ...
ohm       Unit     name name_expression                 ...
ratio     Unit       name name_expression               ...
second    Unit        name name_expression              ...
Name: 0, dtype: object

nsy.u

Hertz     Unit       name name_expression               ...
Farad     Unit       name name_expression               ...
meter     Unit       name name_expression               ...
ohm       Unit     name name_expression                 ...
ratio     Unit       name name_expression               ...
second    Unit        name name_expression              ...
Name: 0, dtype: object

Constant

e_0 = nsy.Constant(
    name="permittivity_vaccum",
    symbol= "\epsilon_0",
    numerical=8.8541878128e-12,
    unit=farad_per_meter
)
e_0

e_0.s

$\displaystyle \epsilon_0$

e_0.n

8.8541878128e-12

e_d = nsy.Constant(
    name="dielectric_permittivity",
    symbol= "\epsilon_d",
    numerical=5,
    unit=nsy.u.ratio
)
e_d

Constants cannot be mutated

e_d.n = 10

Constant cannot be mutated. You cannot set any attribute value. Instantiate a new variable.

Variable

capacitor_plate_separation = nsy.Variable(
    name="capacitor_plate_separation",
    symbol= "d",
    numerical=None,
    unit=nsy.u.meter
)
capacitor_plate_separation

capacitor_plate_separation.s

$\displaystyle d$

capacitor_plate_separation.u

Variables can be mutated

capacitor_plate_separation.n = 1e-6
capacitor_plate_separation.n

1e-06

capacitor_plate_separation.numerical = 3e-5
capacitor_plate_separation.numerical

3e-05

Operate between Value objects

Constants and Variables are value objects.

capacitance_per_plate_cross_sectional_area = e_d / (e_0 * capacitor_plate_separation)
capacitance_per_plate_cross_sectional_area

capacitance_per_plate_cross_sectional_area.se

$\displaystyle \frac{\epsilon_d}{\epsilon_0 d}$

capacitance_per_plate_cross_sectional_area.n

1.8823484456216984e+16

Perform Flexible Class Operations

raw_capacitor_cross_sectional_area = (1e-6) ** 2
raw_capacitor_cross_sectional_area

1e-12

device_capacitance = capacitance_per_plate_cross_sectional_area * raw_capacitor_cross_sectional_area
device_capacitance

device_capacitance.name

''

device_capacitance.se

$\displaystyle \frac{\epsilon_d Ø}{\epsilon_0 d}$

device_capacitance.symbol = "F"
device_capacitance.symbol

$\displaystyle F$

raw_capacitor_cross_sectional_area

1e-12

Example Functions

nsy.sqrt(device_capacitance)

nsy.sinh(device_capacitance)

Future plans

• Extend unit management and verification.
• Create a full constants list, probably even in Excel or as an importable CSV file into Pandas.

Open to contributions.