Uncertainty Module

You might read the user guide introduction of this module.

class maabara.uncertainty.Sheet(equation='0', name='', data=[])

Symbolic propagation of uncertainty

equation : sympy equation string, optional

See set_equation()

name : string, optional

See set_name()

data : list

See set_data()

batch(data, fields, mode='default')

Batch process a set of values.

data : array_like NxM or tuple of array_like

A tuple will be stacked to columns. Columns represent different variables indexed by fields. Rows will be iterated.

fields : string

List of columns fields divided by |. Deviations columns must be suffix by %. Use * to ignore a column form data

mode : {‘default’, ‘ufloat’, or ‘exact’}, optional

Specify return type

out : ndarray

Depending on mode.

‘default’ will return Nx2 array. First column will hold nominal value, second its deviation. ‘exact’ like default but rounded to significant digits. ‘ufloat’ will return Nx1 array of ufloats

It is possible to set constant values or errors before calling batch method. The batch method will automaticly use the values if not given by the data array. See example below.

Retrieve a computation object and set equation

>>> stack = ma.uncertainty.Sheet('a*x**3')

Data array specifies the sets to compute row-by-row

>>> data = [ [0.5, 1. , 0.1],  [0.3, 2. ,0.15] ]    

Batch the data

>>> stack.set_value('a', error=0.05)    # set constant error for a
>>> stack.batch(data, 'a|x|x%')
array([[ 0.5       ,  0.15811388],
       [ 2.4       ,  0.6720119 ]])     # line-by-line result with uncertainty         

Try again with a constant value

>>> stack.set_value('a',1., 0.05)          # define constant a value
>>> stack.batch(data, '*|x|x%', 'ufloat')  # rerun computation ignoring first data column
array([[1.0+/-0.30000000000000004],
       [8.0+/-1.7999999999999998]], dtype=object)

eq(equation, name='')

Alias of set_equation()

get_data(line=False, element=False)

Get data

line : string, optional

Variable name

element : {‘val’, ‘dev’, or ‘tex’}, optional

Return mode

out : mixed

If line and element are False complete data is returned. If found line will be returned as tuple. If element is specified value (val), deviation (dev) or Latex markup (tex) will be returned.

get_result(mode='default')

Get error propagation

mode : {‘default’, ‘exact’, ‘ufloat’, or ‘tex’}

out : mixed

n(name)

Alias of set_name()

p(mode='default', multiply='dot')

Alias of print_result()

print_result(mode='default', multiply='dot')

Print results of symbolic error propagation

mode : {‘default’, or ‘short’}

Specifies printing mode.

multiply : string, optional

Latex markup for multiply symbol. Use None to supress multipliers

out : ufloat

Result

ps()

Alias of set_result() in short mode

reset()

Clear all settings

out : boolean

True on success.

run(equation=False, data=[])

Runs symbolic error propagation by specified data

equation : string, optional

See set_equation()

data : list, optional

See set_data()

out : sympy equation, sympy error_equation, ufloat result

set_data(data)

Set data manually.

data : list of tuples

[ ( string Variable, float Value, float Error = 0, tex = “”) ]

It is more recommend to use set_value() function to manipulate the data.

out : boolean

True on success.

set_equation(equation, name='')

Set equation string

equation : Sympy equation string

Use fundamental functions like sin, exp, sqrt, atan etc. and Rational(a,b) to define a fraction a/b. Allowed variable names are alphanumeric and use _ only. They must be specifed by set_value(). Please be aware that I will be interpreted as imaginary unit.

name : string, optional

See set_name()

out : boolean

True on success.

set_name(name)

Set a name

name : string

Name as Latex markup. It will appear in computation equations.

out : boolean

True on success.

set_value(symbol, value=False, error=False, tex=False)

Set uncertain value

symbol : string

Symbol string used in equation

value : float, optional

Value

error : float, optional

Deviation

tex : string, optional

Latex markup replace string. All symbol occurences will be replaced by this markup.

v(symbol, value=False, error=False, tex=False)

Alias of set_value()