propygate

A simple measurement and uncertainty propagation library

What is it?

This is a Python library dedicated to aid students and professionals in dealing with measurements and uncertainty propagation.

Do you have to do a experimental physics report and forget how to properly estimate sqrt((21.665 ± 0.002) * (16 ± 0.6) ^ (11/12)) or are you just feeling too lazy to do it by hand? Then this may be the right package for you!

Are you a real scientist doing real science stuff? Then this probably isn't the correct library for you. Try using the uncertainties library instead.

How to install it?

The easy way (not working yet...)

pip install propygate

The less easy way

1 - Download the repository

git clone https://github.com/luizsol/propygate.git

2 - Install the package

pip install ./propygate

How to use it?

# We start by importing the Measurement class

from propygate import Measurement

# Now we create some measurements

# This is one way of creating a measurement of 10 ± 0.2
m1 = Measurement(10, error=0.2)

# This is another way
m1 = Measurement(10, 0.2)  # Easy, right?

# Now you can treat this measurement (mostly) as any other numeric variable
# in python

# You can add it to other measurements
m2 = Measurement(4.4, 0.05)
print(m1 + m2)  # 14.4 ± 0.20615528128088306

# Subtract them
print(m1 - m2)  # 5.6 ± 0.20615528128088306

# Divide them
print(m1 / m2)  # 2.2727272727272725 ± 0.05227926003669418

# Multiply them
print(m1 * m2)  # 44.0 ± 1.0121264743103995

# Exponentiate them (but not one by another!)
import math

print(m1 ** (math.pi))  # 1385.1601859244622 ± 87.02961448163397

# You can also use python's default numeric types to make your calculations

print(3.333 + m2)  # 7.7330000000000005 ± 0.05

# But sadly there are some drawbacks to this package, here are them

# Using math standard functions such as sqrt, sin, cos etc. will fail
try:
    print(math.sqrt(m1))
except:
    print('Told ya!')

# For this more advanced stuff you may want to use the uncertainties
# (https://pythonhosted.org/uncertainties/) package

# Here's a tip: instead of using sqrt just exponentiate it to 1/2:

print(m1 ** (1 / 2))  # 3.1622776601683795 ± 0.0316227766016838

Next features

Add tests
Pypi submission
Unit support
Significant figure rule adherence