test_Autocorrelation.py

import pytest
 
from MPCDAnalysis.Autocorrelation import Autocorrelation
from MPCDAnalysis.OnTheFlyStatisticsDDDA import OnTheFlyStatisticsDDDA
 
def test_constructor():
    with pytest.raises(TypeError):
        Autocorrelation(1.0)
 
    with pytest.raises(ValueError):
        Autocorrelation(-1)
 
    Autocorrelation(0)
    Autocorrelation(1)
    Autocorrelation(2)
 
 
def test_getMaxCorrelationTime():
    for N_max in range(0, 5):
        ac = Autocorrelation(N_max)
        assert ac.getMaxCorrelationTime() == N_max
 
 
def test_correlationTimeIsAvailable():
    N_max = 10
 
    def testExceptions(ac):
        for N in range(0, ac.getMaxCorrelationTime() + 1):
            with pytest.raises(TypeError):
                ac.correlationTimeIsAvailable(float(N))
 
        with pytest.raises(ValueError):
            ac.correlationTimeIsAvailable(-1)
 
        with pytest.raises(ValueError):
            ac.correlationTimeIsAvailable(ac.getMaxCorrelationTime() + 1)
 
 
 
    ac = Autocorrelation(N_max)
 
    testExceptions(ac)
 
    for datum in range(0, N_max * 2):
        dataPointsSuppliedSoFar = datum
        for N in range(0, N_max + 1):
            if dataPointsSuppliedSoFar >= N + 1:
                expected = True
            else:
                expected = False
            assert ac.correlationTimeIsAvailable(N) == expected
 
        ac.addDatum(datum)
 
    testExceptions(ac)
 
 
def test_getAutocorrelation():
    import random
    n_max = 100
    N_max = 10
 
    data = [random.uniform(-10, 10) for _ in range(0, n_max)]
 
    ac = Autocorrelation(N_max)
 
    for datumIndex, datum in enumerate(data):
        ac.addDatum(datum)
 
        dataSupplied = data[0 : datumIndex + 1]
        for N in range(0, N_max + 1):
            if not ac.correlationTimeIsAvailable(N):
                continue
 
            ddda = OnTheFlyStatisticsDDDA()
            for i in range(0, len(dataSupplied)):
                if i + N >= len(dataSupplied):
                    break
                ddda.addDatum(dataSupplied[i] * dataSupplied[i + N])
 
            assert ac.getAutocorrelation(N) == ddda
 
 
def test_getAutocorrelation_Vector3DReal():
    import random
    n_max = 100
    N_max = 10
 
    from MPCDAnalysis.Vector3DReal import Vector3DReal
    data = [
        Vector3DReal(
            random.uniform(-10, 10),
            random.uniform(-10, 10),
            random.uniform(-10, 10))
        for _ in range(0, n_max)]
 
    ac = Autocorrelation(N_max)
 
    def mul(v1, v2):
        return v1.dot(v2)
 
    for datumIndex, datum in enumerate(data):
        ac.addDatum(datum, mul)
 
        dataSupplied = data[0 : datumIndex + 1]
        for N in range(0, N_max + 1):
            if not ac.correlationTimeIsAvailable(N):
                continue
 
            ddda = OnTheFlyStatisticsDDDA()
            for i in range(0, len(dataSupplied)):
                if i + N >= len(dataSupplied):
                    break
                ddda.addDatum(dataSupplied[i].dot(dataSupplied[i + N]))
 
            assert ac.getAutocorrelation(N) == ddda