QuantiPhy Eval — Computations with Physical Quantities

Author:: Ken Kundert
Version:: 0.5.0
Released:: 2022-09-02

A companion to QuantiPhy, quantiphy_eval evaluates strings containing simple algebraic expressions that involve quantities. It returns a quantity. For example:

>>> from quantiphy_eval import evaluate

>>> avg_price = evaluate('($1.2M + $1.3M)/2', '$')
>>> print(avg_price)
$1.25M

>>> avg_freq = evaluate('(122.317MHz + 129.349MHz)/2', 'Hz')
>>> print(avg_freq)
125.83 MHz

QuantiPhy Eval is used in networth to allow you to give your estimated values using expressions that include numbers that have units, SI scale factors, and commas. That allows you the convenience of copy-and-pasting your numbers from websites without being forced to reformat them.

With QuantiPhy the units do not survive operations, so you can specify the resolved units using the second argument. In fact, the second argument is passed to QuantiPhy as the model, which allows you to give the return value a name and description along with units, as demonstrated in the next example.

By default QuantiPhy Eval provides no built-in constants. However, you can add your own constants:

>>> from quantiphy import Quantity
>>> from quantiphy_eval import evaluate, initialize
>>> import math

>>> my_constants = dict(
...     k = Quantity('k'),
...     q = Quantity('q'),
...     T = Quantity('25°C', scale='K'),
...     π = Quantity(math.pi),
...     τ = Quantity(math.tau),
... )
>>> initialize(variables=my_constants)

>>> Vt = evaluate('k*T/q', 'Vt V thermal voltage')
>>> print(Vt.render(show_label='f'))
Vt = 25.693 mV — thermal voltage

Alternatively, you can specify the model directly in the text passed to evaluate. Simply append it in the form of a double-quoted string:

>>> Vt = evaluate('k*T/q "Vt V thermal voltage"')
>>> print(Vt.render(show_label='f'))
Vt = 25.693 mV — thermal voltage

You can also use evaluate to assign values to names directly, QuantiPhy Eval remembers these values between calls to evaluate:

>>> f_0 = evaluate('f₀ = 1MHz')
>>> omega_0 = evaluate('ω₀ = τ*f₀ "rads/s"')
>>> print(omega_0.render(show_label=True))
ω₀ = 6.2832 Mrads/s

Similarly, QuantiPhy Eval provides no built-in functions by default, but you can add any you need:

>>> def median(*args):
...    args = sorted(args)
...    l = len(args)
...    m = l//2
...    if l % 2:
...        return args[m]
...    return (args[m] + args[m-1])/2

>>> initialize(functions = dict(median=median))
>>> median_price = evaluate('median($636122, $749151, $706781)', '$')
>>> print(median_price.fixed(show_commas=True))
$706,781

initialize takes three arguments, variables, functions and quantity. Both arguments and functions take dictionaries that overwrite any previously saved values. quantity takes a quantiphy Quantity class. The return value of evaluate will be an object of this class.

rm_commas is a function for removing commas from an expression. This is used if your number contain commas. Simply stripping the commas it would prevent you from using multi-argument functions. However after removing the commas rm_commas also converts semicolons to commas. So the previous example could be rewritten as:

>>> from quantiphy_eval import evaluate, rm_commas

>>> median_price = evaluate(
...     rm_commas('median($636,122; $749,151; $706,781)'),
...     '$',
... )
>>> print(median_price.fixed(show_commas=True))
$706,781

QuantiPhy Eval supports comments. A # and anything that follows it to the end of the line is ignored:

>>> average_price = evaluate(
...     rm_commas('''
...         median(
...             $636,122 +   # Zillow
...             $749,151 +   # Redfin
...             $706,781     # Trulia
...         )/3
...     '''),
...     '$'
... )
>>> print(average_price.fixed(show_commas=True, prec=2, strip_zeros=False))
$697,351.33

Finally, QuantiPhy Eval uses inform.Error for error reporting:

>>> from inform import Error

>>> try:
...     Vt = evaluate('kT/q', 'V')
...     print(Vt)
... except Error as e:
...     print(str(e))
kT: variable unknown.

Releases

Latest development release:

Version: 0.5.0

Released: 2022-09-02

0.5 (2022-09-02):

refactor the project structure
provide qe example, a simple calculator

0.4 (2021-01-27):

Add ability to explicitly specify units (or model) in evaluated string.

0.3 (2020-08-12):

complete re-write, parser now implemented with ply rather than pyparsing.
all built-in constants and functions have been removed.
split evaluate into two: evaluate and initialize.

0.2 (2020-03-06):

rm_commas now converts semicolons to commas
support comments

0.1 (2020-03-05):

Add support for user-defined constants and functions.
add rm_commas function.

0.0 (2020-02-14):

Initial version.