Choosing a Type

quantia has four types. The choice depends on whether your value is scalar or vector, and whether it is exact or uncertain.

	Scalar	Vector
Exact	UnitFloat	UnitArray
Uncertain	ProbUnitFloat	ProbUnitArray

UnitFloat — exact scalar

Use when you have one value that is known precisely:

reservoir_pressure = qu.Q(3500.0, 'psia')
depth              = qu.Q(2450.0, 'm')
temperature        = qu.Q(95.0,   '°C')

UnitArray — exact vector

Use when you have a list of measurements all in the same unit:

layer_depths      = qu.QA([1200.0, 1450.0, 1700.0], 'm')
permeabilities    = qu.QA([12.5, 8.3, 45.2], 'mD')

ProbUnitFloat — uncertain scalar

Use when a value has uncertainty that should propagate through calculations. Internally stores Monte Carlo samples.

with qu.config(n_samples=3000):
    porosity = qu.ProbUnitFloat.triangular(0.12, 0.18, 0.25, '1')
    Bo       = qu.ProbUnitFloat.normal(1.25, 0.05, 'm3_res')

ProbUnitArray — uncertain vector

Use when you have multiple uncertain values of the same quantity, for example a set of layer thicknesses each with their own uncertainty:

with qu.config(n_samples=2000):
    layers = qu.QPA([
        qu.ProbUnitFloat.triangular(8.0,  12.0, 18.0, 'm'),
        qu.ProbUnitFloat.triangular(5.0,   8.0, 14.0, 'm'),
        qu.ProbUnitFloat.triangular(15.0, 20.0, 28.0, 'm'),
    ])

layers.means()       # UnitArray of mean thicknesses
layers.intervals()   # list of (P2.5, P97.5) tuples
layers.intervals(0.8)   # list of (P10, P90) tuples

Mixing types

Exact values can be combined with uncertain values freely. The result is always uncertain:

Vp       = qu.Q(1_000_000.0, 'm3_res')    # exact
porosity = qu.ProbUnitFloat.uniform(...)   # uncertain
Vp_pore  = Vp * porosity                   # → ProbUnitFloat