Users’ Guide


QuantiPhy adds support for quantities to Python. Quantities are little more than a number combined with its units. They are used to represent physical quantities. Your height and weight are both quantities, having both a value and units, and both are important. For example, if I told you that Mariam’s weight was 8, you might assume pounds as the unit of measure if you lived in the US and think Mariam was an infant, or you might assume stones as the units if you live in the UK and assume that she was an adult, or you might assume kilograms if you lived anywhere else and assume she was a small child. The units are very important, and in general it is always best to keep the unit of measure with the number and present the complete value when working with quantities. To do otherwise invites confusion. Just ask NASA. Readers often stumble on numbers without units as they mentally try to determine the units from context. Quantity values should be treated in a manner similar to money, which is also a quantity. Monetary amounts are almost always given with their units (a currency symbol).

Having a single object represent a quantity in a programming language is useful because it binds the units to the number making it more likely that the units will be presented with the number. In addition, quantities from QuantiPhy provide another important benefit. They naturally support the SI scale factors, which for those that are familiar with them are much easier to read and write than the alternatives. The most common SI scale factors are:

T (1012) tera
G (109) giga
M (106) mega
k (103) kilo
m (10-3) milli
μ (10-6) micro
n (10-9) nano
p (10-12) pico
f (10-15) fempto
a (10-18) atto

Numbers with SI scale factors are commonly used in science and engineering to represent physical quantities because it is easy to read and write numbers both large and small. For example, the distance between the atoms in a silicon lattice is roughly 230 pm whereas the distance to the sun is about 150 Gm. Unfortunately, computers do not normally use SI scale factors. Instead, they use E-notation. The two distances would be written as 2.3e-10 m and 1.5e+11 m. Virtually all computer languages such as Python both read and write numbers in E-notation, but none naturally read or write numbers that use SI scale factors, even though SI is an international standard that has been in place for over 50 years and is widely used.

QuantiPhy is an attempt to address both of these deficiencies. It allows quantities to be represented with a single object that allows the complete quantity to be easily read or written as a single unit. It also naturally supports SI scale factors. As such, QuantiPhy allows computers to communicate more naturally with humans, particularly scientists and engineers.


QuantiPhy is a library that adds support to Python for both reading and writing numbers with SI scale factors and units. The primary working construct for QuantiPhy is Quantity, which is a class whose objects hold the number and units that are used to represent a physical quantity. For example, to create a quantity from a string you can use:

>>> from quantiphy import Quantity

>>> distance_to_sun = Quantity('150 Gm')
>>> distance_to_sun.real

>>> distance_to_sun.units

>>> print(distance_to_sun)
150 Gm

Now distance_to_sun contains an object with two values, the number 150000000000.0 and the units ‘m’. The ‘G’ was interpreted as the giga scale factor, which scales 150 by 109.

It is worth considering the alternative for a moment:

>>> d_sol = float('150000000000.0')
>>> print(f'{d_sol} m')
150000000000.0 m

Ignoring the difficulty in writing and reading the number, there is another important difference. The units are placed in the print statement and not kept with the number. This makes the value ambiguous, it clutters the print statement, and it introduces a vulnerability. When coming back and refactoring your code after some time has passed, you might change the units of the number and forget to change the units in the print statement. This is particularly likely if the number is defined far from where it is printed. The result is that erroneous results are printed and is always a risk when two related pieces of information are specified far from one another. QuantiPhy addresses this issue by binding the value and the units into one object.

Quantity is a subclass of float, and so distance_to_sun can be used just like any real number. For example, you can convert the distance to miles using:

>>> distance_in_miles = distance_to_sun / 1609.34
>>> print(distance_in_miles)

When printed or converted to strings quantities naturally use SI scale factors. For example, you can clean up that distance in miles using:

>>> distance_in_miles = Quantity(distance_to_sun / 1609.34, 'miles')
>>> print(distance_in_miles)
93.206 Mmiles

However, you need not explicitly do the conversion yourself. QuantiPhy provides many of the most common conversions for you:

>>> distance_in_miles = Quantity(distance_to_sun, scale='miles')
>>> print(distance_in_miles)
93.206 Mmiles

Specifying Quantities

Normally, creating a Quantity takes one or two arguments. The first is taken to be the value, and the second, if given, is taken to be the model, which is a source of default values.

The first argument: the value

The value may be given as a float, as a string, or as a quantity. The string may be the name of a known constant or it may represent a number. If the string represents a number, it may be in floating point notation (1200.0), in E-notation (ex: 1.2e+3), or use SI scale factors (1.2k). It may also include the units. And like Python in general, the numbers may include underscores to make them easier to read (they are ignored). For example, any of the following ways can be used to specify 1ns:

>>> period = Quantity(1e-9, 's')
>>> print(period)
1 ns

>>> period = Quantity('0.000_000_001 s')
>>> print(period)
1 ns

>>> period = Quantity('1e-9s')
>>> print(period)
1 ns

>>> period = Quantity('1ns')
>>> print(period)
1 ns

>>> period2 = Quantity(period)
>>> print(period2)
1 ns

If given as a string, the value may also be the name of a known constant:

>>> k = Quantity('k')
>>> q = Quantity('q')
>>> print(k, q, sep='\n')
13.806e-24 J/K
160.22e-21 C

The following constants are pre-defined: h, ħ, k, q, c, 0°C, ε₀, μ₀, and Z₀. You may add your own constants.

Currency units ($€¥£₩₺₽₹Ƀ₿Ξ) are a bit different than other units in that they are placed at the front of the quantity.

>>> print(Quantity('$11_200_000'))

>>> print(Quantity(11.2e6, '$'))

When using currency units, if the number has a sign, it should precede the units:

>>> print(Quantity('-$11_200_000'))

>>> print(Quantity(-11.2e6, '$'))

When given as a string, the number may use any of the following scale factors (though you can use the input_sf preference to prune this list if desired):

Q (1030) quetta
R (1027) ronna
Y (1024) yotta
Z (1021) zetta
E (1018) exa
P (1015) peta
T (1012) tera
G (109) giga
M (106) mega
k (103) kilo
_ (1)
c (10-2) centi
m (10-3) milli
u (10-6) micro (ASCII)
µ (10-6) micro (unicode micro)
μ (10-6) micro (unicode Greek mu)
n (10-9) nano
p (10-12) pico
f (10-15) fempto
a (10-18) atto
z (10-21) zepto
y (10-24) yocto
r (10-27) ronto
q (10-30) quecto

In addition, the units must start with a letter or any of these characters: °ÅΩƱΩ℧¢$€¥£₩₺₽₹Ƀ₿șΞΔ%√, and may be followed by those characters (except %) or digits or any of these characters: -^/()·⁻⁰¹²³⁴⁵⁶⁷⁸⁹. Thus, any of the following would be accepted as units: Ohms, V/A, J-s, m/s^2, H/(m-s), Ω, %, m·s⁻², V/√Hz.

When specifying the value as a string you may also give a name and description, and if you do they become available as the attributes name and desc. This conversion is under the control of the assign_rec preference. The default version of assign_rec accepts either ‘=’ or ‘:’ to separate the name from the value, and either ‘—’, ‘–’, ‘#’, or ‘//’ to separate the value from the description if a description is given. Thus, by default QuantiPhy recognizes specifications of the following forms:

<name> = <value>
<name> = <value> — <description>
<name> = <value> -- <description>
<name> = <value> # <description>
<name> = <value> // <description>
<name>: <value>
<name>: <value> — <description>
<name>: <value> -- <description>
<name>: <value> # <description>
<name>: <value> // <description>

For example:

>>> period = Quantity('Tclk = 10ns -- clock period')
>>> print(f'{} = {period}  # {period.desc}')
Tclk = 10 ns  # clock period

The second argument: the model

If you only specify a real number for the value, then the units, name, and description do not get values. Even if given as a string or quantity, the value may not contain these extra attributes. This is where the second argument, the model, helps. It may be another quantity or it may be a string. Any attributes that are not provided by the first argument are taken from the second if available. If the second argument is a string, it is split. If it contains one value, that value is taken to be the units, if it contains two, those values are taken to be the name and units, and it it contains more than two, the remaining values are taken to be the description. If the model is a quantity, only the units are inherited. For example:

>>> out_period = Quantity(10*period, period)
>>> print(out_period)
100 ns

>>> freq = Quantity(100e6, 'Hz')
>>> print(freq)
100 MHz

>>> freq = Quantity(100e6, 'Fin Hz')
>>> print(f'{} = {freq}')
Fin = 100 MHz

>>> freq = Quantity(100e6, 'Fin Hz input frequency')
>>> print(f'{} = {freq}{freq.desc}')
Fin = 100 MHz — input frequency

If the model contains units, those units are only used if the value does not have units. The same is true for the description. For example:

>>> h = Quantity('18in', 'm')
>>> print(h)
18 in

The remaining arguments

Any arguments beyond the first two must be given as named arguments.

If you need to override the name, units or the description given in either the value or the model, you can do so by specifying them with corresponding named arguments. For example:

>>> out_period = Quantity(
...     10*period, period, name='output period',
...     desc='period at output of frequency divider'
... )
>>> print(f'{} = {out_period}{out_period.desc}')
output period = 100 ns — period at output of frequency divider

In this the value is 10*period, which is a float and so has no name, units, or description attributes, but the model is period that has all three attributes, but the name name and description, coming from a quantity, are ignored. Instead, they are specified explicitly using the name and desc arguments.

Specifying binary as True allows you to use the binary scale factors. The binary scale factors are Ki, Mi, Gi, Ti, Pi, Ei, Zi, and Yi. Unlike the normal scale factors, you cannot use a lower case k in Ki. Also, input_sf is ignored. The normal recognizers are used if none of the binary scale factors are found.

>>> bytes = Quantity('1 KiB', binary=True)
>>> print(bytes)
1.024 kB

You can also specify scale and ignore_sf as named arguments. scale allows you to scale the value or convert it to different units. It is described in a bit. ignore_sf indicates that any scale factors should be ignored. This is one way of handling units whose name starts with a scale factor character. For example:

>>> x = Quantity('1m')                                  # unitless value
>>> print(x, x.real, x.units, sep=', ')
1m, 0.001,

>>> l = Quantity('1m', ignore_sf=True)                  # length in meters
>>> print(l, l.real, l.units, sep=', ')
1 m, 1.0, m

>>> d = Quantity('1m', units = 'mile', ignore_sf=True)  # distance in miles
>>> print(d, d.real, d.units, sep=', ')
1 mile, 1.0, mile

>>> t = Quantity('1m', units = 'min', ignore_sf=True)   # duration in minutes
>>> print(t, t.real, t.units, sep=', ')
1 min, 1.0, min

Finally, you can also specify conversion parameters using params. These values are ignored by QuantiPhy except that they are made available to any UnitConversion conversion functions as a way of implementing parametrized conversions.

Quantity attributes

You can overwrite Quantity attributes to override the units, name, or description.

>>> out_period = Quantity(10*period)
>>> out_period.units = 's'
>>> = 'output period'
>>> out_period.desc = 'period at output of frequency divider'
>>> print(f'{} = {out_period}{out_period.desc}')
output period = 100 ns — period at output of frequency divider

In addition, you can also override the preferences with attributes:

>>> out_period.spacer = ''
>>> print(out_period)

Scaling When Creating a Quantity

Quantities tend to be used primarily when reading and writing numbers, and less often when processing numbers. Often data comes in an undesirable form. For example, imagine data that has been normalized to kilograms but the numbers themselves have neither units or scale factors. QuantiPhy allows you to scale the number and assign the units when creating the quantity:

>>> mass = Quantity('2.529', scale=1000, units='g')
>>> print(mass)
2.529 kg

In this case the value is given in kilograms, and is converted to the base units of grams by multiplying the given value by 1000. You always want to convert to base units (units with no scale factor) when creating a Quantity. This can also be expressed as follows:

>>> mass = Quantity('2.529', scale=(1000, 'g'))
>>> print(mass)
2.529 kg

You can also specify a function to do the conversion, which is helpful when the conversion is not linear:

>>> def from_dB(value, units=''):
...     return 10**(value/20), value.units[2:]

>>> Quantity('-100 dBV', scale=from_dB)
Quantity('10 uV')


Since version 2.18 the first argument, in this case value, is guaranteed to be a Quantity that contains both the units and any parameters needed during the conversion. As such, the second argument, units, is not longer needed and will eventually be removed.

The conversion can also often occur if you simply state the units you wish the quantity to have:

>>> Tboil = Quantity('212 °F', scale='K')
>>> print(Tboil)
373.15 K

or if you employ a subclass of Quantity that has units:

>>> class Kelvin(Quantity):
...     units = 'K'

>>> Tboil = Kelvin('212 °F')
>>> print(Tboil)
373.15 K

This assumes that the initial value is specified with units. If not, you need to provide them for these mechanisms to work.

>>> Tboil = Quantity('212', '°F', scale='K')
>>> print(Tboil)
373.15 K

To do this conversion, QuantiPhy examines the given units (°F) and the desired units (K) and chooses the appropriate converter. No scaling is done if the given units are equivalent as the desired units. Thus you can use the scaling mechanism to convert a collection of data with mixed units to values with consistent units. For example:

>>> weights = '''
...     240 lbs
...     230 lb
...     100 kg
...     210
... '''.strip().split('\n')
>>> for weight in weights:
...     w = Quantity(weight, 'lb', scale='lb')
...     print(w)
240 lb
230 lb
220.46 lb
210 lb

To perform these conversions QuantiPhy uses predefined relationships between pairs of units. These relationships are defined using Unit Converters.

When using unit conversions it is important to only convert to units without scale factors when creating a quantity. For example, it is better to convert to ‘g’ rather than ‘kg’. Otherwise, if the desired units used when creating a quantity includes a scale factor, it is easy to end up with two scale factors when converting the number to a string (ex: 1 mkg or one milli-kilo-gram).

Here is another example that uses quantity scaling. Imagine that a table is being read that gives temperature versus time, but the temperature is given in °F and the time is given in minutes and neither are given with units. Assume that for the purpose of later analysis it is desirable for the values be converted to the more natural units of Kelvin and seconds:

>>> rawdata = '0 450, 10 400, 20 360'
>>> data = []
>>> for pair in rawdata.split(','):
...     time, temp = pair.split()
...     time = Quantity(time, 'min', scale='s')
...     temp = Quantity(temp, '°F', scale='K')
...     data += [(time, temp)]

>>> for time, temp in data:
...     print(f'{time:9q} {temp:9q}')
      0 s  505.37 K
    600 s  477.59 K
   1.2 ks  455.37 K

Creating a Quantity by Scaling an Existing Quantity

The Quantity.scale() method scales the value of a quantity and then uses the new value to create a new Quantity. For example:

>>> import math

>>> h_line = Quantity('1420.405751786 MHz')
>>> sagan = h_line.scale(math.pi)
>>> sagan2 = sagan.scale(2)
>>> print(sagan, sagan2, sep='\n')
4.4623 GHz
8.9247 GHz

>>> print(repr(h_line))
Quantity('1.420405751786 GHz')

>>> print(repr(sagan))
Quantity('4.462336274928 GHz')

Any value that can be passed to the scale argument for Quantity or Quantity.render() can be passed to the scale method. Specifically, the following types are accepted:

float or Quantity

The argument scales the underlying value (a new quantity is returned whose value equals the underlying quantity multiplied by scale). In this case the scale is assumed unitless (any units are ignored) and so the units of the new quantity are the same as those of the underlying quantity.


The argument consists of two values. Tthe first value, a float, is treated as a scale factor. The the second value, a string, is taken to be the units of the new quantity.


The function takes two arguments, the value to be scaled and its units. The value is guaranteed to be a Quantity that includes the units, so the second argument is redundant and will eventually be deprecated. The function returns two values, the value and units of the new value.


The argument is taken to the be desired units. This value along with the units of the underlying quantity are used to select a known unit conversion, which is applied to create the new value.

>>> Tboil_C = Tboil.scale('C')
>>> print(Tboil_C)
100 C

Creating a Quantity by Adding to an Existing Quantity

The Quantity.add() method adds a contribution to the value of a quantity and then uses the sum to create a new Quantity. For example:

>>> import math

>>> total = Quantity(0, '$')
>>> for contribution in ['1.23', '4.56', '7.89']:
...     total = total.add(contribution)
>>> print(total)

The argument to add can be a quantity, a real number, or a string.

When adding quantities, the units of the quantity should match. You can enforce this by adding check_units=True. If the dimension of your quantities match but not the units, you can often use Quantity.scale() to get the units right:

>>> m1 = Quantity('1kg')
>>> m2 = Quantity('1lb')
>>> m3 = m1.add(m2.scale('g'), check_units=True)
>>> print(m3)
1.4536 kg

Accessing Quantity Values

There are a variety of ways of accessing the value of a quantity. If you are just interested in its numeric value, you access it with:

>>> h_line.real

>>> float(h_line)

Or you can simply use a quantity in the same way that you would use any real number, meaning that you can use it in expressions and it evaluates to its numeric value:

>>> second_sagan_freq = 2 * math.pi * h_line
>>> print(second_sagan_freq)

>>> sagan2 = Quantity(second_sagan_freq, h_line)
>>> print(sagan2)
8.9247 GHz

>>> type(h_line)
<class 'quantiphy.quantiphy.Quantity'>

>>> type(second_sagan_freq)
<class 'float'>

>>> type(sagan2)
<class 'quantiphy.quantiphy.Quantity'>

Notice that when performing arithmetic operations on quantities the units are completely ignored and do not propagate in any way to the newly computed result.

If you are interested in the units of a quantity, you can use:

>>> h_line.units

Or you can access both the value and the units, either as a tuple or in a string:

>>> h_line.as_tuple()
(1420405751.786, 'Hz')

>>> str(h_line)
'1.4204 GHz'

SI scale factors are used by default when converting numbers to strings. The following scale factors could be used: QRYZEPTGMkc%munpfazyrq, though by default % is treated as a unit rather than a scale factor. You need to activate % in input_sf for it to be treated as a scale factor.

Only the scale factors listed in the output_sf preference are actually used, and by default that is set to TGMkmunpfa, which avoids the more uncommon scale factors. You can set output_sf to Quantity.all_sf to output all known scale factors except c or %, which are never used in output.

The Quantity.render() method allows you to control the process of converting a quantity to a string. For example:

>>> h_line.render()
'1.4204 GHz'

>>> h_line.render(form='eng')
'1.4204e9 Hz'

>>> h_line.render(show_units=False)

>>> h_line.render(form='eng', show_units=False)

>>> h_line.render(prec=6)
'1.420406 GHz'

>>> h_line.render(form='fixed', prec=2)
'1420405751.79 Hz'

>>> bytes.render(form='binary')
'1 KiB'

>>> k.render(negligible=1e-12)
'0 J/K'

show_label allows you to display the name and description of the quantity when rendering. If show_label is False, the quantity is not labeled with the name or description. Otherwise the quantity is labeled under the control of the show_label value and the show_desc, label_fmt and label_fmt_full preferences (described further in Preferences and Quantity.set_prefs()). If show_label is ‘a’ (for abbreviated) or if the quantity has no description, label_fmt is used to label the quantity with its name. If show_label is ‘f’ (for full), label_fmt_full is used to label the quantity with its name and description. Otherwise label_fmt_full is used if show_desc is True and label_fmt otherwise.

>>> freq.render(show_label=True)
'Fin = 100 MHz'

>>> freq.render(show_label='f')
'Fin = 100 MHz — input frequency'

>>> Quantity.set_prefs(show_desc=True)
>>> freq.render(show_label=True)
'Fin = 100 MHz — input frequency'

>>> freq.render(show_label='a')
'Fin = 100 MHz'

You can also access the full precision of the quantity:

>>> h_line.render(prec='full')
'1.420405751786 GHz'

>>> h_line.render(form='eng', prec='full')
'1.420405751786e9 Hz'

Full precision implies whatever precision was used when specifying the quantity if it was specified as a string and if the keep_components preference is True. Otherwise a fixed number of digits, specified in the full_prec preference, is used (default=12). Generally one uses ‘full’ when generating output that is intended to be read by a machine without loss of precision.

An alternative to render is Quantity.fixed(). It converts the quantity to a string in fixed-point format:

>>> total = Quantity('$11.2M')
>>> print(total.fixed(prec=2, show_commas=True, strip_zeros=False))

You can also use Quantity.render() to produce a fixed format, but it does not support all of the options available with fixed:

>>> print(total.render(form='fixed', prec=2))

Another alternative to render is Quantity.binary(). It converts the quantity to a string that uses binary scale factors:

>>> mem = Quantity(17_179_869_184, 'B', name='physical memory')
>>> print(mem.binary())
16 GiB

Alternatively you can also use render to render strings with binary prefixes:

>>> print(mem.render(form='binary'))
16 GiB

Scaling When Rendering a Quantity

Once it comes time to output quantities from your program, you may again may be constrained in the way the numbers must be presented. QuantiPhy also allows you to scale the values as you render them to strings. In this case, the value of the quantity itself remains unchanged. For example, imagine having a quantity in grams and wanting to present it in either kilograms or in pounds:

>>> m = Quantity('2529 g')
>>> print("mass (kg): {}".format(m.render(show_units=False, scale=0.001)))
mass (kg): 2.529

>>> print(m.render(scale=(0.0022046, 'lb'), form='fixed'))
5.5754 lb

As before, functions can also be used to do the conversion. Here is an example where that comes in handy: a logarithmic conversion to dBV is performed.

>>> import math
>>> def to_dB(value, units=''):
...     return 20*math.log10(value), 'dB'+value.units

>>> T = Quantity('100mV')
>>> print(T.render(scale=to_dB))
-20 dBV


Since version 2.18 the first argument, in this case value, is guaranteed to be a Quantity that contains both the units and any parameters needed during the conversion. As such, the second argument, units, is not longer needed and will eventually be removed.

Finally, you can also use either the built-in converters or the converters you created to do the conversion simply based on the units:

>>> print(m.render(scale='lb'))
5.5755 lb

In an earlier example the units of time and temperature data were converted to normal SI units. Presumably this makes processing easier. Now, when producing the output, the units can be converted back to the original units if desired:

>>> for time, temp in data:
...     print("{:<7} {}".format(time.render(scale='min'), temp.render(scale='°F')))
0 min   450 °F
10 min  400 °F
20 min  360 °F

String Formatting

Quantities can be passed into the string format method:

>>> print('{}'.format(h_line))
1.4204 GHz

>>> print('{:s}'.format(h_line))
1.4204 GHz

In these cases the preferences for SI scale factors, units, and precision are honored.

Specifying the format

You can override the precision as part of the format specification

>>> print('{:.6}'.format(h_line))
1.420406 GHz

You can also specify the width and alignment. Quantiphy follows the Python convention of right justifying numbers by default.

>>> print({:16.6}»'.format(h_line))
«    1.420406 GHz»

>>> print({:<16.6}»'.format(h_line))
«1.420406 GHz    »

>>> print({:>16.6}»'.format(h_line))
«    1.420406 GHz»

>>> print({:^16.6}»'.format(h_line))
«  1.420406 GHz  »

The general form of the format specifiers supported by quantities is:

format_spec ::=  [align][#][width][,][.precision][type][scale]

align specifies the alignment using one of the following characters:




Right justification.


Left justification.


Center justification.

The hash (#) is a literal hash that when present indicates that trailing zeros and radix should not be stripped from the fractional part of the number.

width is a literal integer that specifies the minimum width of the string.

The comma (,) is a literal comma that when present indicates that commas should be added to the whole part of the mantissa, every three digits.

precision is a literal integer that specifies the precision.

And finally, type specifies which form should be used when formatting the value. The choices include:




Use default formatting options.


Use default formatting options.


Format using SI scale factors and show the units.


Format using SI scale factors but do not show the units.


Format using fixed-point notation and show the units.


Format using exponent notation but do not show the units.


Format using fixed-point notation but do not show the units.


Format using binary prefixes while showing the units.


Format using fixed-point or exponential notation, whichever is shorter, but do not show the units.


Only include the units.


Only include the name.


Only include the description.

You can capitalize any of the format characters that output the value of the quantity (any of ‘sqrpefg’, but not ‘und’). If you do, the label will also be included.

These format specifiers are generally included in format strings. However, in addition, Quantitphy provides the Quantity.format() method that converts a quantity to a string based on a naked format string. For example:

>>> print(h_line.format('.6q'))
1.420406 GHz

Any format specification that is not recognized by QuantiPhy is simply passed on to the underlying float. For example:

>>> print(f'TOTAL: {total:+#,.2f}')
TOTAL: +11,200,000.00

>>> with Quantity.prefs(input_sf='%'):
...     growth = Quantity('23.7%')
>>> print(f'growth = {growth:.0%}')
growth = 24%


Here is an example of these format types:

>>> h_line = Quantity('f = 1420.405751786 MHz — hydrogen line')
>>> for f in 'sSpPqQrRbBeEfFgGund':
...     print(f + ':', h_line.format(f))
s: 1.4204 GHz
S: f = 1.4204 GHz — hydrogen line
p: 1420405751.786 Hz
P: f = 1420405751.786 Hz — hydrogen line
q: 1.4204 GHz
Q: f = 1.4204 GHz — hydrogen line
r: 1.4204G
R: f = 1.4204G — hydrogen line
b: 1.3229 GiHz
B: f = 1.3229 GiHz — hydrogen line
e: 1.4204e+09
E: f = 1.4204e+09 — hydrogen line
f: 1420405751.786
F: f = 1420405751.786 — hydrogen line
g: 1.4204e+09
G: f = 1.4204e+09 — hydrogen line
u: Hz
n: f
d: hydrogen line

The ‘q’ type specifier is used to explicitly indicate that both the number and the units are desired and that SI scale factors should be used, regardless of the current preferences.

>>> print('{:.6q}'.format(h_line))
1.420406 GHz

Alternately, ‘r’ can be used to indicate just the number represented using SI scale factors is desired, and the units should not be included.

>>> print('{:r}'.format(h_line))

The opposite can be achieved using ‘p’, which includes the units without SI scale factors:

>>> print('{:p}'.format(h_line))
1420405751.786 Hz

The ‘p’ format is often used with ‘#’ to format currency values:

>>> print('{:#.2p}'.format(total))

>>> print('{:#,.2p}'.format(total))

The ‘b’ format is used to render number with binary scale factors:

>>> print('{:b}'.format(mem))
16 GiB

>>> print('{:B}'.format(mem))
physical memory = 16 GiB

You can also use the traditional floating point format type specifiers:

>>> print('{:f}'.format(h_line))

>>> print('{:e}'.format(h_line))

>>> print('{:g}'.format(h_line))

Use ‘u’ to indicate that only the units are desired:

>>> print('{:u}'.format(h_line))

Access the name or description of the quantity using ‘n’ and ‘d’.

>>> print('{:n}'.format(freq))

>>> print('{:d}'.format(freq))
input frequency

Using the upper case versions of the format codes that print the numerical value of the quantity (SQRFEG) indicates that the quantity should be labeled with its name and perhaps its description (as if the show_label preference were set). They are under the control of the show_desc, label_fmt and label_fmt_full preferences (described further in Preferences and Quantity.set_prefs()).

If show_desc is False or the quantity does not have a description, then label_fmt is used to add the labeling.

>>> Quantity.set_prefs(show_desc=False)
>>> trise = Quantity('10ns', name='trise')

>>> print('{:S}'.format(trise))
trise = 10 ns

>>> print('{:Q}'.format(trise))
trise = 10 ns

>>> print('{:R}'.format(trise))
trise = 10n

>>> print('{:F}'.format(trise))
trise = 0

>>> print('{:E}'.format(trise))
trise = 1e-08

>>> print('{:G}'.format(trise))
trise = 1e-08

>>> print('{0:n} = {0:q} ({0:d})'.format(freq))
Fin = 100 MHz (input frequency)

>>> print('{:S}'.format(freq))
Fin = 100 MHz

If show_desc is True and the quantity has a description, then label_fmt_full is used if the quantity has a description.

>>> Quantity.set_prefs(show_desc=True)

>>> print('{:S}'.format(trise))
trise = 10 ns

>>> print('{:S}'.format(freq))
Fin = 100 MHz — input frequency

Scaling while formatting

Finally, you can add units after the format code, which causes the number to be scaled to those units if the transformation represents a known unit conversion. In this case the format code must be specified (use ‘s’ rather than ‘’).

>>> Tboil = Quantity('Boiling point = 100 °C')
>>> print('{:S°F}'.format(Tboil))
Boiling point = 212 °F

>>> eff_channel_length = Quantity('leff = 14nm')
>>> print(f'{eff_channel_length:}')
leff = 140 Å

>>> print(f'{mem:bb}')
128 Gib

This feature can be used to simplify the conversion of the time and temperature information back into the original units:

>>> for time, temp in data:
...     print(f'{time:<7smin} {temp:s°F}')
0 min   450 °F
10 min  400 °F
20 min  360 °F

You can add a scale factor to the units, in which case the number will be scaled accordingly:

>>> for p in range(1, 5):
...     bytes = Quantity(256**p, 'B')
...     print(f"An {8*p} bit bus addresses {bytes:,pkB}.")
An 8 bit bus addresses 0.256 kB.
An 16 bit bus addresses 65.536 kB.
An 24 bit bus addresses 16,777.216 kB.
An 32 bit bus addresses 4,294,967.296 kB.

Generally you should only specify base units when using a format that renders with scale factors as otherwise you could see two scale factors on the same number. For example, if the q format was used in the above example, the last address space would be rendered as 4.295 MkB.

Ambiguity of Scale Factors and Units

By default, QuantiPhy treats both the scale factor and the units as being optional. With the scale factor being optional, the meaning of some specifications can be ambiguous. For example, ‘1m’ may represent 1 milli or it may represent 1 meter. Similarly, ‘1meter’ my represent 1 meter or 1 milli-eter. In this case QuantiPhy gives preference to the scale factor, so ‘1m’ normally converts to 1e-3. To allow you to avoid this ambiguity, QuantiPhy accepts ‘_’ as the unity scale factor. In this way ‘1_m’ is unambiguously 1 meter. You can instruct QuantiPhy to output ‘_’ as the unity scale factor by specifying the unity_sf argument to Quantity.set_prefs():

>>> Quantity.set_prefs(unity_sf='_', spacer='')
>>> l = Quantity(1, 'm')
>>> print(l)

This is often a good way to go if you are outputting numbers intended to be read unambiguously or by both people and machines.

If you need to interpret numbers that have units and are known not to have scale factors, you can specify the ignore_sf preference:

>>> Quantity.set_prefs(ignore_sf=True, unity_sf='', spacer=' ')
>>> l = Quantity('1000m')
>>> l.as_tuple()
(1000.0, 'm')

>>> print(l)
1 km

>>> Quantity.set_prefs(ignore_sf=False)
>>> l = Quantity('1000m')
>>> l.as_tuple()
(1.0, '')

If there are scale factors that you know you will never use, you can instruct QuantiPhy to interpret a specific set and ignore the rest using the input_sf preference.

>>> Quantity.set_prefs(input_sf='GMk')
>>> l = Quantity('1000m')
>>> l.as_tuple()
(1000.0, 'm')

>>> print(l)
1 km

Specifying input_sf=None causes QuantiPhy to again accept all known scale factors.

>>> Quantity.set_prefs(input_sf=None)
>>> l = Quantity('1000m')
>>> l.as_tuple()
(1.0, '')

Alternatively, you can specify the units you wish to use whose leading character is a scale factor. Once known, these units no longer confuse QuantiPhy. These units can be specified as a list or as a string. If specified as a string the string is split to form the list. Specifying the known units replaces any existing known units.

>>> d1 = Quantity('1 au')     # astronomical unit
>>> d2 = Quantity('1000 pc')  # parsec
>>> p = Quantity('138 Pa')    # Pascal
>>> print(d1.render(form='eng'), d2, p, sep='\n')
1e-18 u
1 nc
138e15 a

>>> Quantity.set_prefs(known_units='au pc Pa')
>>> d1 = Quantity('1 au')
>>> d2 = Quantity('1000 pc')
>>> p = Quantity('138 Pa')
>>> print(d1.render(form='eng'), d2, p, sep='\n')
1 au
1 kpc
138 Pa

This same issue comes up for temperature quantities when given in Kelvin. There are again several ways to handle this. First you can specify the acceptable input scale factors leaving out ‘K’, ex. input_sf = ‘TGMkmunpfa’, or:

>>> Quantity.set_prefs(input_sf=Quantity.get_pref('input_sf').replace('K', ''))
>>> temp = Quantity('100K')
>>> print(temp.as_tuple())
(100.0, 'K')

>>> temp = Quantity('100k')
>>> print(temp.as_tuple())
(100000.0, '')

>>> temp = Quantity('100k', 'K')
>>> print(temp.as_tuple())
(100000.0, 'K')

Alternatively, you can specify ‘K’ as one of the known units. Finally, if you know exactly when you will be converting a temperature to a quantity, you can specify ignore_sf for that specific conversion. The effect is the same either way, ‘K’ is interpreted as a unit rather than a scale factor.

The same techniques would be used to handle volumes in cubic centimeters:

>>> vol = Quantity('10 cc')
>>> print(vol.as_tuple())
(0.1, 'c')
>>> with Quantity.prefs(input_sf=Quantity.get_pref('input_sf').replace('c', '')):
...     vol = Quantity('10 cc')
>>> print(vol.as_tuple())
(10.0, 'cc')
>>> with Quantity.prefs(known_units='cc'):
...     vol = Quantity('100 cc')
>>> print(vol.as_tuple())
(100.0, 'cc')

Percentages are a special case. QuantiPhy can treat the % character as either a unit or a scale factor (0.01). By default it is treated as a unit:

>>> tolerance = Quantity('10%')
>>> change = Quantity('10%Δ')
>>> print(tolerance.as_tuple(), change.as_tuple(),)
(10.0, '%') (10.0, '%Δ')

If, however, you add % as a known scale factor, it then acts as a scale factor.

>>> with Quantity.prefs(input_sf = Quantity.get_pref('input_sf') + '%'):
...     tolerance = Quantity('10%')
...     change = Quantity('10%Δ')
...     print(tolerance.as_tuple(), change.as_tuple(),)
(0.1, '') (0.1, 'Δ')

In general you cannot simply add to the list of known scale factors. The % character is an exception as QuantiPhy knows about it but disables it by default.

Subclassing Quantity

You can subclass Quantity to make it easier to create a particular type of quantity, or to create quantities with particular qualities. The following example demonstrates both. It creates a subclass for dollars that both sets the units and the display preferences. Any Quantity preference (see Quantity.set_prefs()) may be given as an attribute. Display preferences for currencies are often very different from what you would want from physical quantities:

>>> class Dollars(Quantity):
...     units = '$'
...     form = 'fixed'
...     prec = 2
...     strip_zeros = False
...     show_commas = True

>>> cost = Dollars(100_000)
>>> print(cost)

This example creates a special class for bytes.

>>> class Bytes(Quantity):
...     units = 'B'
...     form = 'binary'
...     accept_binary = True

>>> memory = Bytes('64KiB')
>>> print(memory)
64 KiB

Here, two classes are created for voltage and current, each with their own perspective on what values should be considered negligible.

>>> class Voltage(Quantity):
...     units = 'V'
...     negligible = 1e-6

>>> class Current(Quantity):
...     units = 'A'
...     negligible = 1e-12

>>> Vout = Voltage(1e-9)
>>> Ileak = Current(1e-9)
>>> print(f"Vout = {Vout}, Ileak = {Ileak}.")
Vout = 0 V, Ileak = 1 nA.

Lastly, this example creates a special class for temperatures. It disallows use of ‘K’ as a scale factor to avoid confusion with Kelvin units.

>>> class Temperature(Quantity):
...     units = 'K'
...     input_sf = Quantity.get_pref('input_sf').replace('K', '')

>>> Tcore = Temperature('15M')
>>> Tphoto = Temperature('5.3k')
>>> Tcmb = Temperature('3.18K')
>>> print(Tcore, Tphoto, Tcmb, sep='\n')
15 MK
5.3 kK
3.18 K

Scaling with Subclasses

Special scaling rules come into play if the units attribute is present on a Quantity class. In such a case you can specify the class as an argument to a scaling operation. For example:

>>> class Grams(Quantity):
...     units = 'g'

>>> class Pounds(Quantity):
...     units = 'lbs'

>>> wt = Pounds(10)
>>> mass = wt.scale(Grams)

>>> print(mass, repr(mass), sep='\n')
4.5359 kg
Grams('4.5359237 kg')

>>> print(wt.render(scale=Grams))
4.5359 kg

Notice that use of Grams with the Quantity.scale() method resulted in a return value of type Grams. This does not naturally occur if you scale using scale factors or units:

>>> mass = wt.scale('g')
>>> print(mass, repr(mass), sep='\n')
4.5359 kg
Quantity('4.5359237 kg')

In this case you can replicate the previous behavior by adding Grams as an argument to the conversion:

>>> mass = wt.scale('g', cls=Grams)
>>> print(mass, repr(mass), sep='\n')
4.5359 kg
Grams('4.5359237 kg')

Scaling Upon Subclass Creation

When creating quantities using a subclass, a conversion automatically occurs if both the subclass and the value have units. The conversion converts the given units to those expected by the class. For example:

>>> class Seconds(Quantity):
...     units = 's'

>>> ttl = Seconds('2 days')
>>> print(ttl)
172.8 ks

If you also specify a scale argument, that conversion occurs before the result is converted to the units of the class:

>>> class Days(Quantity):
...     units = 'days'

>>> expires = Days('48 hr', scale='s')
>>> print(expires)
2 days

Adding the scale argument is handy because QuantiPhy does not provide a built-in direct conversion between hours and days. In this case two conversions occur, from hours to seconds, as a result of the scale request, and from seconds to days, to convert to the units expected by the class.

Unit Converters

The UnitConversion class defines conversion relationships between pairs of units, which saves you the trouble of having to remember the actual conversion factors. Once defined, a relationship is available anywhere in QuantiPhy where a unit conversion can occur. For example:

>>> from quantiphy import Quantity, UnitConversion

>>> m_smoot = UnitConversion('m', 'smoots', 1.7)

>>> length_of_harvard_bridge = Quantity('364.4 smoots')
>>> print(length_of_harvard_bridge.render(scale='m', prec=1))
620 m

This is a linear conversion. This unit conversion says, when converting smoots to m, multiply by 1.7. When going the other way, divide by 1.7.

You can also specify units with a scale factor when scaling a number. For example, you can explicitly direct that the length of the bridge should be output in kilometers using:

>>> print(f"{length_of_harvard_bridge:.2pkm}")
0.62 km

QuantiPhy* provides a collection of built-in converters for common units:

base units

related units

C °C

K, F °F, R °R


C °C, F °F, R °R


micron, Å angstrom, mi mile miles, ft feet, in inch inches


oz, lb lbs


sec second seconds, min minute minutes, hour hours hr, day days



BTC btc Ƀ ₿

sat sats ș

The conversions can occur between a pair of units, one from the first column and one from the second. They do not occur when both units are only in the second column. So for example, it is possible to convert between g and lbs, but not between oz and lb. However, if you notice, the units in the second column are grouped using commas. A set of units within commas are considered equivalent, meaning that there are multiple names for the same underlying unit. For example, in, inch, and inches are all considered equivalent. You can convert between equivalent units even though both are found in either the first or second columns.

UnitConversion supports linear conversions (slope only), affine conversions (slope and intercept) nonlinear conversions, parameterized conversions (conversions with extra parameters) and dynamic conversions (convertions that change over time). Here are some examples:

>>> def from_dB(dB):
...     return 10**(dB/20)

>>> def to_dB(v):
...     return 20*math.log10(v)

>>> m_inch = UnitConversion('m', 'in inch inches', 0.0254)  # linear
>>> C_F = UnitConversion('C °C', 'F °F', 5/9, -32*5/9)      # affine
>>> _dB = UnitConversion('', 'dB', from_dB, to_dB)          # nonlinear

>>> print(Quantity('12 in', scale='m'))
304.8 mm

>>> print(Quantity('100 °C', scale='°F'))
212 °F

>>> print(Quantity('100', scale='dB'))
40 dB

One thing to be aware of with affine conversions like °C to °F: they are suitable for converting absolute temperatures but not temperature differences. One way around this is to add another conversion specifically for differences:

>>> dC_F = UnitConversion('ΔC Δ°C', 'ΔF Δ°F', 5/9)
>>> print(Quantity('100 Δ°C', scale='Δ°F'))
180 Δ°F

Notice that the scaling functions used here differ from those described previously in that they only take one argument and return one value. The units are not included in either then argument list or the return value.

Also notice that the return value of UnitConversion was not used in the examples above. It is enough to simply create the UnitConversion for it to be available to Quantity. So, it is normal to not capture the return value of UnitConversion. However, there are a few things you can do with the return value. First you can convert it to a string to get a description of the relationship. This is largely used as a sanity check:

>>> print(C_F)
C ← 0.5555555555555556*F + -17.778

In addition, you can use it to directly perform conversions:

>>> temp_F = C_F.convert(0, '°C', '°F')
>>> print(temp_F)
32 °F

>>> temp_C = C_F.convert(32, '°F', '°C')
>>> print(temp_C)
0 °C

Finally, you can pre-define multiple conversions between the same pairs of units, and activate the one you currently wish to use. This can be useful with conversions that change over time. For example

>>> btc_usd_2017_peak = UnitConversion('USD $', 'BTC Ƀ', 19870.62)
>>> btc_usd_2021_peak = UnitConversion('USD $', 'BTC Ƀ', 68978.64)

>>> print(Quantity("5 BTC", scale='$'))

>>> btc_usd_2017_peak.activate()
>>> print(Quantity("5 BTC", scale='$'))

>>> btc_usd_2021_peak.activate()
>>> print(Quantity("5 BTC", scale='$'))

Defining a conversion between the same pair of units acts to conceal an earlier definition, but the previous definition can be restored using activate().

Parametrized Unit Converters

Occasionally you might encounter conversion that requires one or more extra parameters. For example, to convert between concentration and molarity in solutions requires the atomic weight of the solute. These extra parameters can be passed in when creating a quantity and then are available to the desired conversion. For example:

>> @UnitConversion.fixture
>> def from_molarity(M, mw):
..     return M * mw

>> @UnitConversion.fixture
>> def to_molarity(g_L, mw):
..     return g_L / mw

>> mol_conv = UnitConversion('g/L', 'M', from_molarity, to_molarity)

>> KCl_conc = Quantity('1.2 mg/L', params=74.55)
>> print(f"{KCl_conc:qM}")
16.097 uM

For more information on defining unit converters, see UnitConversion. For more information on parametrized unit converters, see UnitConversion.fixture(). For example of real-time dynamic conversions, see Dynamic Unit Conversions.

Scale Factor Conversions

In the preceding sections it was shown that you can use the scaling features of QuantiPhy to convert between units using only the name of the units. When doing so the relationship between the units must be known, and UnitConversion is used to specify the relationship. However, it is also possible to perform simple scale factor conversions without changing the units. This case is specified in a manner similar to a unit conversion, but in this case both the from-units and the to-units are the same, and it is not necessary to define a UnitConversion. For example, imagine printing a table of bit-rates where the rates are held in bps but are expected to be displayed in Mbps:

>>> rates = [155.52e6, 622.08e6, 2.48832e9, 9.95328e9, 39.81312e9]
>>> rates = [Quantity(r, 'bps') for r in rates]
>>> for r in rates:
...     print(f"{r:>14,.2pMbps}")
   155.52 Mbps
   622.08 Mbps
 2,488.32 Mbps
 9,953.28 Mbps
39,813.12 Mbps

You can also do the inverse; convert simple numbers given in Mbps to quantities in bps:

>>> rates = [155.52, 622.08, 2488.32, 9953.28, 39813.12]
>>> rates = [Quantity(r, 'Mbps', scale='bps') for r in rates]
>>> for r in rates:
...     print(r.as_tuple())
(155520000.0, 'bps')
(622080000.0, 'bps')
(2488320000.0, 'bps')
(9953280000.0, 'bps')
(39813120000.0, 'bps')

Quantity Functions

It is sometimes handy to convert directly to and from real values rather than converting to Quantity objects and holding them. Generally it is preferred to key a value and its units together, but as said before, the primary use of QuantiPhy is inputting and outputting numbers. If you are not inputting and outputting the same numbers, it may not be worth even the small overhead of a Quantity object. In that case, you can use quantity functions to convert directly to and from real values. If you wish to use QuantiPhy to convert to a simple float, use as_real(). It takes the same arguments as a Quantity, but returns a float rather than a Quantity:

>>> from quantiphy import as_real
>>> print(as_real('10 mL'))

It is common to use Scale Factor Conversions to scale the result to the desired size:

>>> print(as_real('10 mL', scale='uL'))

as_tuple() is similar except it returns both the value and the units as a tuple:

>>> from quantiphy import as_tuple
>>> print(as_tuple('10 mL'))
(0.01, 'L')

>>> print(as_tuple('10 mL', scale='uL'))
(10000.0, 'uL')

Finally, you can use render(), fixed(), and binary() to convert a real value and units into a string. Besides the value and the units, the these functions the same arguments as Quantity.render(), Quantity.fixed(), and Quantity.binary().

>>> from quantiphy import render, fixed, binary
>>> print(render(1e-6, 'L'))
1 uL

>>> print(fixed(1e7, '$', show_commas=True, strip_zeros=False, prec=2))

>>> print(binary(2**32, 'B'))
4 GiB

Physical Constants

QuantiPhy has several built-in constants that are available by specifying their name to the Quantity class. The following quantities are built in:


MKS value

CGS value



6.626070040e-34 J-s

6.626070040e-27 erg-s

Plank’s constant

hbar, ħ

1.054571800e-34 J-s

1.054571800e-27 erg-s

Reduced Plank’s constant


1.38064852e-23 J/K

1.38064852e-16 erg/K

Boltzmann’s constant


1.6021766208e-19 C

4.80320425e-10 Fr

Elementary charge


2.99792458e8 m/s

2.99792458e8 m/s

Speed of light

0C, 0°C

273.15 K

273.15 K

0 Celsius

eps0, ε₀

8.854187817e-12 F/m

Permittivity of free space

mu0, μ₀

4e-7π H/m

Permeability of free space

Z0, Z₀

376.730313461 Ohms

Characteristic impedance of free space

Constants are given in base units (g, m, etc.) rather than the natural units for the unit system (kg, cm, etc.). For example, when using the CGS unit system, the speed of light is given as 300Mm/s (rather than 30Gcm/s).

As shown, these constants are partitioned into two unit systems: mks and cgs. Only those constants that are associated with the active unit system and those that are not associated with any unit system are available when creating a new quantity. You can activate a unit system using set_unit_system(). Doing so deactivates the previous system. By default, the mks system is active.

You can create your own constants and unit systems using add_constant():

>>> from quantiphy import Quantity, add_constant
>>> add_constant(Quantity("λₕ: 211.061140539mm // wavelength of hydrogen line"))

>>> hy_wavelength = Quantity('λₕ')
>>> print(hy_wavelength.render(show_label=True))
λₕ = 211.06 mm — wavelength of hydrogen line

In this case is the name given in the quantity is used when creating the constant. You can also specify an alias as an argument to add_constant.

>>> add_constant(
...     Quantity("λₕ = 211.061140539mm # wavelength of hydrogen line"),
...     alias='lambda h'
... )

>>> hy_wavelength = Quantity('lambda h')
>>> print(hy_wavelength.render(show_label=True))
λₕ = 211.06 mm — wavelength of hydrogen line

It is not necessary to specify both the name and the alias, one is sufficient; the constant is accessible using either. Notice that the alias does not actually become part of the constant, it is only used for looking up the constant.

By default, user defined constants are not associated with a unit system, meaning that they are always available regardless of which unit system is being used. However, when creating a constant you can specify one or more unit systems for the constant. You need not limit yourself to the predefined mks and cgs unit systems. You can specify multiple unit systems either by specifying a list of strings for the unit systems, or by specifying one string that would contain more than one name once split.

>>> from quantiphy import Quantity, add_constant, set_unit_system

>>> add_constant(Quantity(4.80320427e-10, 'Fr'), 'q', 'esu gaussian')
>>> add_constant(Quantity(1.602176487e-20, 'abC'), alias='q', unit_systems='emu')
>>> q_mks = Quantity('q')
>>> set_unit_system('cgs')
>>> q_cgs = Quantity('q')
>>> set_unit_system('esu')
>>> q_esu = Quantity('q')
>>> set_unit_system('gaussian')
>>> q_gaussian = Quantity('q')
>>> set_unit_system('emu')
>>> q_emu = Quantity('q')
>>> set_unit_system('mks')
>>> print(q_mks, q_cgs, q_esu, q_gaussian, q_emu, sep='\n')
160.22e-21 C
480.32 pFr
480.32 pFr
480.32 pFr
16.022e-21 abC


QuantiPhy supports a wide variety of preferences that control its behavior. For example, when rendering quantities you can control the number of digits used (prec), whether SI scale factors are used (form), whether the units are included (show_units), etc. Similar preferences also control the conversion of strings into quantities, which can help disambiguate whether a suffix represents a scale factor or a unit. The list of available preferences and their descriptions are given in the description of the Quantity.set_prefs() method.

To set a preference, use the Quantity.set_prefs() class method. You can set more than one preference at once:

>>> Quantity.set_prefs(prec=6, map_sf={'u': 'μ'})

This statements tells QuantiPhy to use 7 digits (the prec plus 1) and to output μ rather u for the 10-6 scale factor.

Setting preferences to None returns them to their default values:

>>> Quantity.set_prefs(prec=None, map_sf=None)

The preferences are changed on the class itself, meaning that they affect any instance of that class regardless of whether they were instantiated before or after the preferences were set. If you would like to have more than one set of preferences, then you should subclass Quantity. For example, imagine a situation where you have different types of quantities that would naturally want different preferences:

>>> class Temperature(Quantity):
...     units = 'C'
>>> Temperature.set_prefs(prec=1, known_units='K', spacer='')

>>> class Frequency(Quantity):
...     units = 'Hz'
>>> Frequency.set_prefs(prec=5, spacer='')

>>> frequencies = []
>>> for each in '-25.3 999987.7, 25.1  1000207.1, 74.9  1001782.3'.split(','):
...     temp, freq = each.split()
...     frequencies.append((Temperature(temp),  Frequency(freq)))

>>> for temp, freq in frequencies:
...     print(f'{temp:4}  {freq}')
-25C  999.988kHz
 25C  1.00021MHz
 75C  1.00178MHz

In this example, a subclass is created that is intended to report in concentrations.

>>> class Concentration(Quantity):
...     pass
>>> Concentration.set_prefs(
...     map_sf = dict(u=' PPM', n= ' PPB', p=' PPT'),
...     show_label = True,
... )

>>> pollutants = dict(CO=5, SO2=20, NO2=0.10)
>>> concentrations = [Concentration(v, scale=1e-6, name=k) for k, v in pollutants.items()]
>>> for each in concentrations:
...     print(each)
CO = 5 PPM
SO2 = 20 PPM
NO2 = 100 PPB

Alternately, you can simply set the preferences as attributes when creating the sublclasses. For example:

>>> class Dollars(Quantity):
...     units = '$'
...     prec = 2
...     form = 'fixed'
...     show_commas = True
...     minus = Quantity.minus_sign
...     strip_zeros = False

When a subclass is created, the preferences active in the main class are copied into the subclass. Subsequent changes to the preferences in the main class do not affect the subclass.

You can also go the other way and override the preferences on a specific quantity.

>>> print(hy_wavelength)
211.06 mm

>>> hy_wavelength.show_label = True
>>> print(hy_wavelength)
λₕ = 211.06 mm — wavelength of hydrogen line

This is often the way to go with quantities that have logarithmic units such as decibels (dB) or shannons (Sh) (or the related bit, digits, nats, hartleys, etc.). In these cases use of SI scale factors is often undesired.

>>> gain = Quantity(0.25, 'dB')
>>> print(gain)
250 mdB

>>> gain.form = 'fixed'
>>> print(gain)
0.25 dB

To retrieve a preference, use the Quantity.get_pref() class method. This is useful with known_units. Normally setting known_units overrides the existing units. You can simply add more with:

>>> Quantity.set_prefs(known_units=Quantity.get_pref('known_units') + ['K'])

A variation on Quantity.set_prefs() is Quantity.prefs(). It is basically the same, except that it is meant to work with Python’s with statement to temporarily override preferences:

>>> with Quantity.prefs(form='fixed', show_units=False, prec=2):
...     for time, temp in data:
...         print(f"{time:<7} {temp}")
0       505.37
600     477.59
1200    455.37

>>> print(f"Final temperature = {data[-1][1]} @ {data[-1][0]}.")
Final temperature = 455.37 K @ 1.2 ks.

Notice that the specified preferences only affected the table, not the final printed values, which were rendered outside the with statement.

If you are using QuantiPhy in a large package with multiple modules and more than one includes Quantity, you may find that the preferences are not shared between the modules. This occurs because each module gets its own independent version of Quantity. To work around this issue you would create your own module that imports from QuantiPhy. Each of the packages’ modules then import from your new module rather than directly from QuantiPhy. For example, consider creating a local module named

from quantiphy import *
import locale

# Base preferences
loc_conv = locale.localeconv()
radix = loc_conv['decimal_point']
comma = loc_conv['thousands_sep']
Quantity.set_prefs(radix=radix, comma=comma, known_units='K')

# Alternate preference sets
preferences = dict(
    user = dict(
        # assumes a user is reading values on a terminal with fixed-width font
        form = 'si',
        map_sf = Quantity.map_sf_to_greek,
        prec = 4,
        spacer = ' ',
        strip_radix = True,
        minus = Quantity.minus_sign,
        show_units = True,
    sphinx = dict(
        # assumes values are to be rendered with a variable-with font by Sphinx
        form = 'si',
        map_sf = Quantity.map_sf_to_sci_notation,
        prec = 4,
        spacer = Quantity.narrow_non_breaking_space,
        minus = Quantity.minus_sign,
        strip_radix = True,
        show_units = True,
    code_with_si = dict(
        # assumes values are to be rendered to code that accepts sf but not units
        form = 'sia',
        map_sf = None,
        prec = 'full',
        spacer = '',
        minus = '-'.minus_sign,
        strip_radix = 'cover',  # assures quantities are always treated as reals
    code_without_si = dict(
        # assumes values are to be rendered to code that does not accept sf or units
        form = 'eng',
        map_sf = None,
        prec = 'full',
        spacer = '',
        minus = '-'.minus_sign,
        strip_radix = 'cover',  # assures quantities are always treated as reals

def set_quantity_defaults(choice):


Now, in the other modules, you would simply import from quantity rather than quantiphy:

from quantity import Quantity, QuantiPhyError, set_quantity_defaults

Then, if you change the preferences using set_quantity_defaults from one module, the preferences are changed for all modules.


Quantiphy provides 7 preferences that help with localization: radix, comma, plus, minus, inf, nan, and spacer.


The decimal point; generally . or ,.


The thousands separator; generally ,, ., _ or a narrow non-breaking space.


QuantitPhy does not use plus signs when rendering quantities either on the mantissa or the exponent. But it will accept this symbol as a plus signs when converting strings to quantities.


The symbol used to indicate a negative number; generally - or . This symbol is also accepted as a minus signs when converting strings to quantities.


The symbol or word that signifies infinity; generally inf or .


The symbol or word that indicates a NaN or Not-a-Number; generally NaN or nan.


The character used to separate the mantissa from trailing units, or scale factor combined with units: generally `` `` or Quantity.narrow_non_breaking_space. spacer does not affect how strings are converted quantities, where the spacer is optional and may ether be a space, a non-breaking space, a thin space, or a narrow non-breaking space.

By default QuantiPhy uses ., ,, +, -, inf, nan and `` `` as the defaults. These are all simple ASCII characters. They work as expected for the numbers normally used in programming, such as -5.17e+06.

Both radix and comma affect the way stings are converted to quantities and they way quantities are rendered. When interpreting a string as a number, QuantiPhy first strips the comma character from the string and then replaces the radix character with ..

If you prefer to use , for your radix, you generally have two choices. With the first, radix is set to , and comma to .. This allows you to properly read and write numbers like €100.000.000,00 but misinterpretes a number if it uses . as the radix.

>>> Quantity.set_prefs(radix=',', comma='.')
>>> q1 = Quantity('€100.000,00')
>>> q2 = Quantity('€100000.00')
>>> print(q1, q2, sep='\n')

With the second, radix is set to , and comma to ‘’. This allows both , and . to be used as the radix, so €100,000 and €100.000 have the same value. However, it fails for numbers that use . as the thousands separator.

>>> Quantity.set_prefs(radix=',', comma='')
>>> q1 = Quantity('€100,000')
>>> q2 = Quantity('€100.000')
>>> print(q1, q2, sep='\n')

You can automatically adapt to local conventions using the Python locale package:

>>> from quantiphy import Quantity
>>> import locale

>>> loc_conv = locale.localeconv()
>>> radix = loc_conv['decimal_point']
>>> comma = loc_conv['thousands_sep']
>>> Quantity.set_prefs(radix=radix, comma=comma)

>>> q = Quantity('€100.000')
>>> print(q)

>>> print(f"radix is '{radix}'\ncomma is '{comma}'")
radix is '.'
comma is ''

You can convert from one convention to the other by changing radix and comma on the fly:

>>> with Quantity.prefs(radix=',', comma='.'):
...     q = Quantity('€100.000.000,00')
>>> with Quantity.prefs(radix='.', comma=','):
...     print(f'{q:#,.2p}')

Formatting Tabular Data

When creating tables it is often desirable to align the decimal points of the numbers, and perhaps align the units. You can use the number_fmt to arrange this. number_fmt is a format string that if specified is used to convert the components of a number into the final number. You can control the widths and alignments of the components to implement specific arrangements. number_fmt is passed to the string format function with named arguments: whole, frac and units, which contains the integer part of the number, the fractional part including the decimal point, and the units including the scale factor. More information about the content of the components can be found in Quantity.set_prefs().

For example, you can align the decimal point and units of a column of numbers like this:

>>> lengths = [
...     Quantity(l)
...     for l in '1mm, 10mm, 100mm, 1.234mm, 12.34mm, 123.4mm'.split(',')
... ]

>>> with Quantity.prefs(number_fmt='{whole:>3}{frac:<4} {units}'):
...     for l in lengths:
...         print(l)
  1     mm
 10     mm
100     mm
  1.234 mm
 12.34  mm
123.4   mm

You can also give a function as the value for number_fmt rather than a string. It would be called with whole, frac and units as arguments given in that order. The function is expected to return the assembled number as a string. For example:

>>> def fmt_num(whole, frac, units):
...     return '{mantissa:<5} {units}'.format(mantissa=whole+frac, units=units)

>>> with Quantity.prefs(number_fmt=fmt_num):
...     for l in lengths:
...         print(l)
1     mm
10    mm
100   mm
1.234 mm
12.34 mm
123.4 mm

If there are multiple columns it might be necessary to apply a different format to each column. In this case, it often makes sense to create a subclass of Quantity for each column that requires distinct formatting:

>>> def format_temperature(whole, frac, units):
...     return '{:>5} {:<5}'.format(whole+frac, units)

>>> class Temperature(Quantity):
...     units = 'C'
>>> Temperature.set_prefs(
...     prec = 1, known_units = 'K', number_fmt = format_temperature
... )

>>> class Frequency(Quantity):
...     units = 'Hz'
>>> Frequency.set_prefs(prec=5, number_fmt = '{whole:>3}{frac:<6} {units}')

>>> frequencies = []
>>> for each in '-25.3 999987.7, 25.1 1000207.1, 74.9 1001782.3'.split(','):
...     temp, freq = each.split()
...     frequencies.append((Temperature(temp),  Frequency(freq)))

>>> for temp, freq in frequencies:
...     print(temp, freq)
  -25 C     999.988   kHz
   25 C       1.00021 MHz
   75 C       1.00178 MHz

Extract Quantities

It is possible to put a collection of quantities in a text string and then use the Quantity.extract() method to parse the quantities and return them in a dictionary. For example:

>>> design_parameters = '''
...     Fref (fₒ) = 156 MHz  — Reference frequency
...     Kdet = 88.3 uA       — Gain of phase detector
...     Kvco = 9.07 GHz/V    — Gain of VCO
... '''
>>> quantities = Quantity.extract(design_parameters)

>>> Quantity.set_prefs(
...     label_fmt = '{n} = {v}',
...     label_fmt_full = '{V:<18}  # {d}',
...     show_label = 'f',
... )
>>> for k, q in quantities.items():
...     print(f'{k}: {q}')
Fref: fₒ = 156 MHz        # Reference frequency
Kdet: Kdet = 88.3 uA      # Gain of phase detector
Kvco: Kvco = 9.07 GHz/V   # Gain of VCO

The string is processed one line at a time and may contain any number of quantity definitions. Blank lines are ignored. Each non-blank line is passed through assign_rec to determine if it is recognized as an assignment. If it is recognized, the assign_rec named fields (name, qname, val, and desc) are used when creating the quantity. The default recognizer allows you to separate the name from the value with either ‘=’ or ‘:’. It allows you to separate the value from the description using ‘—’, ‘–’, ‘//’, or ‘#’. These substrings are also used to introduce comments, so you could start a line with ‘#’ and it would be treated as a comment. If the line is not recognized, then it is ignored.

In this example, the first line is nonconforming and so is ignored. The second Kvdo line is a comment, the comment character and anything beyond is ignored. Finally, empty lines are ignored.

>>> design_parameters = '''
...     PLL Design Parameters
...     Fref = 156 MHz      — Reference frequency
...     Kdet = 88.3 uA      — Gain of phase detector
...     Kvco = 9.07 GHz/V   — Gain of VCO
...     // Kvco = 5 GHz/V     — Gain of VCO
...     N = 128             — Divide ratio
...     Fout = N*Fref "Hz"  — Output Frequency
... '''
>>> globals().update(Quantity.extract(design_parameters))

>>> print(f'{Fref:S}\n{Kdet:S}\n{Kvco:S}\n{N:S}\n{Fout:}')
Fref = 156 MHz      # Reference frequency
Kdet = 88.3 uA      # Gain of phase detector
Kvco = 9.07 GHz/V   # Gain of VCO
N = 128             # Divide ratio
Fout = 19.968 GHz   # Output Frequency

In this case the output of the Quantity.extract() call is fed into globals().update() so as to add the quantities into the module namespace, making the quantities accessible as local variables. This is an example of how simulation scripts could be written. The system and simulation parameters would be gathered together at the top into a multiline string, which would then be read and loaded into the local name space. It allows you to quickly give a complete description of a collection of parameters when the goal is to put something together quickly in an expressive manner. Another example of this ideas is shown a bit further down where the module docstring is used to contain the quantity definitions.

Here is an example that uses this feature to read parameters from a file. This is basically the same idea as above, except the design parameters are kept in a separate file. It also subclasses Quantity to create a version that displays the name and description by default.

>>> from quantiphy import Quantity, InvalidNumber
>>> from inform import os_error, fatal, display

>>> class VerboseQuantity(Quantity):
...    show_label = 'f'
...    label_fmt = '{n} = {v}'
...    label_fmt_full = '{V:<18}{d}'

>>> filename = '.parameters'
>>> try:
...     with open(filename) as f:
...         globals().update(VerboseQuantity.extract(
... except OSError as e:
...     fatal(os_error(e))
... except InvalidNumber as e:
...     fatal(e, culprit=filename)

>>> print(Fref, Kdet, Kvco, N, Fout, sep='\n')
Fref = 156 MHz     — Reference frequency
Kdet = 88.3 uA     — Gain of phase detector (Imax)
Kvco = 9.07 GHz/V  — Gain of VCO
N = 128            — Divide ratio
Fout = 19.968 GHz  — Output Frequency

With Quantity.extract() the values of quantities can be given as a expression that contains previously defined quantities (or physical constants or select mathematical constants (pi, tau, π, or τ). You can follow an expression with a string to give the units. Finally, you can use the predefined argument to pass in a dictionary of named values that can be used in your expressions. For example:

#!/usr/bin/env python3
>>> __doc__ = """
... Simulates a second-order ΔΣ modulator with the following parameter values:
...     Fclk = Fxtal/4 "Hz"                  — clock frequency
...     Fin = 200kHz                         — input frequency
...     Vin = 950mV                          — input voltage amplitude (peak)
...     gain1 = 0.5V/V                       — gain of first integrator
...     gain2 = 0.5V/V                       — gain of second integrator
...     Vmax = 1V                            — quantizer maximum input voltage
...     Vmin = -1V                           — quantizer minimum input voltage
...     levels = 5                           — quantizer output levels
...     Tstop = 2/Fin "s"                    — simulation stop time
...     Tstart = -1/Fin 's'                  — initial transient interval (discarded)
...     file_name = 'out.wave'               — output filename
...     sim_name = f'{Fclk:q} ΔΣ Modulator'  — simulation name
... The values given above are used in the simulation; no further
... modification of the code given below is required when changing
... these parameters.
... """

>>> from quantiphy import Quantity

>>> Fxtal = Quantity('200 MHz')
>>> parameters = Quantity.extract(__doc__, predefined=dict(Fxtal=Fxtal))
>>> globals().update(parameters)

>>> with Quantity.prefs(
...     label_fmt = '{n} = {v}',
...     label_fmt_full = '{V:<18} ­— {d}',
...     show_label = 'f',
... ):
...     print('Simulation parameters:')
...     for k, v in parameters.items():
...         try:
...             print(f'    {v:Q}')
...         except ValueError:
...             print(f'    {k} = {v!s}')
Simulation parameters:
    Fclk = 50 MHz      ­— clock frequency
    Fin = 200 kHz      ­— input frequency
    Vin = 950 mV       ­— input voltage amplitude (peak)
    gain1 = 500 mV/V   ­— gain of first integrator
    gain2 = 500 mV/V   ­— gain of second integrator
    Vmax = 1 V         ­— quantizer maximum input voltage
    Vmin = -1 V        ­— quantizer minimum input voltage
    levels = 5         ­— quantizer output levels
    Tstop = 10 us      ­— simulation stop time
    Tstart = -5 us     ­— initial transient interval (discarded)
    file_name = out.wave
    sim_name = 50 MHz ΔΣ Modulator

Notice in this case the parameters were specified and read out of the docstring at the top of the file. In this way, the parameters become very easy to set and the documentation is always up to date. Ignore the fact that the docstring is assigned to __doc__. That was a hack that was needed to make the example executable from within the documentation.

Translating Quantities

Quantity.all_from_conv_fmt() recognizes conventionally formatted numbers and quantities embedded in text and reformats them using Quantity.render(). This is an difficult task in general, and so some constraints are placed on the values to make them easier to distinguish. Specifically, the units, if given, must be simple and immediately adjacent to the number. Units are simple if they only consist of letters and underscores. The characters °, Å, Ω and Ʊ are also allowed. So ‘47e3Ohms’, ‘50_Ohms’ and ‘1.0e+12Ω’ are recognized as quantities, but ‘50 Ohms’ and ‘12m/s’ are not.

Besides the text to be translated, all_from_conv_fmt() takes the same arguments as render(), though they must be given as named arguments.

>>> test_results = '''
... Applying stimulus @ 2.00500000e-04s: V(in) = 5.000000e-01V.
... Pass @ 3.00500000e-04s: V(out): expected=2.00000000e+00V, measured=1.99999965e+00V, diff=3.46117130e-07V.
... '''.strip()

>>> Quantity.set_prefs(spacer='')
>>> translated = Quantity.all_from_conv_fmt(test_results)
>>> print(translated)
Applying stimulus @ 200.5us: V(in) = 500mV.
Pass @ 300.5us: V(out): expected=2V, measured=2V, diff=346.12nV.

Quantity.all_from_si_fmt() is similar, except that it recognizes quantities formatted with either a scale factor or units and ignores plain numbers. Again, units are expected to be simple and adjacent to their number.

>>> Quantity.set_prefs(spacer='')
>>> translated_back = Quantity.all_from_si_fmt(translated, form='eng')
>>> print(translated_back)
Applying stimulus @ 200.5e-6s: V(in) = 500e-3V.
Pass @ 300.5e-6s: V(out): expected=2V, measured=2V, diff=346.12e-9V.

Notice in the translations the quantities lost resolution. This is avoided if you use ‘full’ precision:

>>> translated = Quantity.all_from_conv_fmt(test_results, prec='full')
>>> print(translated)
Applying stimulus @ 200.5us: V(in) = 500mV.
Pass @ 300.5us: V(out): expected=2V, measured=1.99999965V, diff=346.11713nV.


You can determine whether a value is equivalent to that of a quantity using Quantity.is_close(). The value may be given as a quantity, a real number, or a string. The two values need not be identical, they just need to be close to be deemed equivalent. The reltol and abstol preferences are used to determine if they are close.

>>> h_line.is_close(h_line)

>>> h_line.is_close(h_line + 1)

>>> h_line.is_close(h_line + 1e4)

Quantity.is_close() returns true if the units match and if:

abs(a - b) <= max(reltol * max(abs(a), abs(b)), abstol)

where a and b represent other and the numeric value of the underlying quantity.

By default, is_close() looks at the both the value and the units if the argument has units. In this way if you compare two quantities with different units, the is_close() test will always fail if their units differ. This behavior can be overridden by specifying check_units.

>>> Quantity('$10').is_close('10 USD')

>>> Quantity('$10').is_close('10 USD', check_units=False)

Negligible Values

QuantiPhy can round small values to zero when rendering them, which can help to reduce visual clutter. You can specify the size of a negligible value as a preference using Quantity.set_prefs() or Quantity.prefs(), or you can specify it locally using Quantity.render(). Any quantity whose absolute value is smaller than the specified value is rendered as zero with the underlying value remaining unchanged.

>>> from quantiphy import Quantity
>>> from math import exp

>>> Vt = 0.025852
>>> def cond(v):
...     return Quantity(1e-27 * exp(v/Vt)/Vt, 'Ʊ')

>>> Quantity.set_prefs(prec=2)
>>> for i in range(11):
...     v = Quantity(i/5, 'V')
...     print(f'{v:>6}: {cond(v):>10}, {v:>26}: {cond(v).render(negligible=1e-3):>10}')
   0 V: 38.7e-27 Ʊ,                        0 V:        0 Ʊ
200 mV: 88.6e-24 Ʊ,                     200 mV:        0 Ʊ
400 mV:  203e-21 Ʊ,                     400 mV:        0 Ʊ
600 mV:     465 aƱ,                     600 mV:        0 Ʊ
800 mV:    1.06 pƱ,                     800 mV:        0 Ʊ
   1 V:    2.44 nƱ,                        1 V:        0 Ʊ
 1.2 V:    5.58 uƱ,                      1.2 V:        0 Ʊ
 1.4 V:    12.8 mƱ,                      1.4 V:    12.8 mƱ
 1.6 V:     29.3 Ʊ,                      1.6 V:     29.3 Ʊ
 1.8 V:      67 kƱ,                      1.8 V:      67 kƱ
   2 V:     153 MƱ,                        2 V:     153 MƱ

Exceptional Values

QuantiPhy supports NaN (not a number) and infinite values:

>>> inf = Quantity('inf Hz')
>>> print(inf)
inf Hz

>>> nan = Quantity('NaN Hz')
>>> print(nan)
NaN Hz

You can test whether the value of the quantity is infinite or is not-a-number using Quantity.is_infinite() or Quantity.is_nan(). These method return a rendered value for the number without units if they are true and None otherwise:

>>> h_line.is_infinite()

>>> inf.is_infinite()

>>> h_line.is_nan()

>>> nan.is_nan()

The rendered value is affected by the inf and nan preferences or attributes:

>>> inf.inf = '∞'
>>> inf.is_infinite()


The way exceptions are defined in QuantiPhy has changed. Initially, the standard Python exceptions were used to indicate errors. For example, a ValueError was raised by Quantity if it were passed a string it cannot convert into a number. Now, a variety of QuantiPhy specific exceptions are used to indicate specific errors. However, these exceptions subclass the corresponding Python error for compatibility with existing code. It is recommended that new code catch the QuantiPhy specific exceptions rather than the generic Python exceptions as their use will be deprecated in the future.

QuantiPhy employs the following exceptions:


Subclass of QuantiPhyError and ValueError. Used by add_constant().

Raised when the value is either not an instance of Quantity or a string that can be converted to a quantity.


Subclass of QuantiPhyError and ValueError. Used by Quantity constructor.

Raised when comma and radix preference is the same.


Subclass of QuantiPhyError and TypeError. Used by Quantity.add().

Raised when the units of contribution do not match those of underlying quantity.


Subclass of QuantiPhyError, ValueError, and TypeError. Used by Quantity().

Raised if the value given could not be converted to a number.


Subclass of QuantiPhyError and KeyError. Used by Quantity().

The assign_rec preference is expected to be a regular expression that defines one or more named fields, one of which must be val. This exception is raised when the current value of assign_rec does not satisfy this requirement.


Subclass of QuantiPhyError and NameError. Used by add_constant().

Raised when alias was not specified and no name was available from value.


Subclass of QuantiPhyError and KeyError.

Used by UnitConversion.convert(), Quantity(), Quantity.scale(), Quantity.render(), Quantity.fixed(), Quantity.format(), Quantity.binary(), as_real(), as_tuple(), render(), fixed(), and binary().

Raised when a unit conversion was requested and there is no corresponding unit converter.


Subclass of QuantiPhyError and KeyError. Used by Quantity.render(), Quantity.fixed(), and Quantity.format().

The label_fmt and label_fmt_full are expected to be format strings that may interpolate certain named arguments. The valid named arguments are n for name, v for value, and d for description. This exception is raised when some other name is used for an interpolated argument.


Subclass of QuantiPhyError and KeyError. Used by Quantity.set_prefs(), Quantity.get_pref(), and Quantity.prefs().

Raised when the name given for a preference is unknown.


Subclass of QuantiPhyError and ValueError. Used by Quantity(), Quantity.set_prefs(), or Quantity.prefs().

The input_sf preference gives the list of scale factors that should be accepted. This exception is raised if input_sf contains an unknown scale factor.


Subclass of QuantiPhyError and KeyError. Used by set_unit_system().

Raised when the name given does not correspond to a known unit system.

QuantiPhy defines a common base exception, QuantiPhyError, that all specific exceptions derive from. This allows you to simplify your exception handling if you are not interested in distinguishing between the specific errors:

>>> from quantiphy import Quantity, QuantiPhyError

>>> try:
...     q = Quantity('tweed')
... except QuantiPhyError as e:
...     print(str(e))
'tweed': not a valid number.

The alternative would be to catch each error individually:

>>> from quantiphy import (
...     Quantity, InvalidNumber, UnknownScaleFactor,
...     UnknownConversion, InvalidRecognizer
... )

>>> try:
...     q = Quantity('tweed')
... except (InvalidNumber, UnknownScaleFactor, UnknownConversion, InvalidRecognizer) as e:
...     print(str(e))
'tweed': not a valid number.

QuantiPhy provides uniform access methods for its exceptions. You can access all the unnamed arguments passed to the exception using the args attribute, you can access the named arguments using kwargs, and you can create a customized message that incorporates the arguments using QuantiPhyError.render() method. The argument to render is a format string that can access both the unnamed and named arguments:

>>> try:
...     q = Quantity('tweed')
... except InvalidNumber as e:
...     print(e.render('{}: no es un número valido.'))
... except UnknownScaleFactor as e:
...     print(e.render('factor de escala desconocido.'))
... except UnknownConversion as e:
...     if 'to_units' in e.kwargs:
...         if 'from_units' in e.kwargs:
...             template = 'incapaz de convertir entre {} y {}'
...         else:
...             template = 'incapaz de convertir a {}'
...     else:
...         template = 'incapaz de convertir de {}'
...     print(e.render(template))
... except InvalidRecognizer as e:
...     print(e.render("el reconocedor no contiene la clave 'val'"))
tweed: no es un número valido.

Alternately, if you wish to globally replace the default QuantiPhy error messages, you can simply override the _template attribute for the exceptions:

>>> InvalidNumber._template = '{!r}: no es un número valido.'
>>> UnknownScaleFactor._template = 'factor de escala desconocido.'
>>> UnknownConversion._template = (
...     'incapaz de convertir entre ‘{to_units}’ y ‘{from_units}’',
...     'incapaz de convertir a ‘{to_units}’',
...     'incapaz de convertir de ‘{from_units}’',
... )
>>> InvalidRecognizer._template = "el reconocedor no contiene la clave ‘val’"

>>> try:
...     q = Quantity('tweed')
... except QuantiPhyError as e:
...     print(e.render())
'tweed': no es un número valido.

As shown in UnknownConversion, _template may be given as a tuple of format strings, in which case the first one for which all arguments are available is used.