def sampleSize(data, proportion):
denominator = power(margin_error(data, sample_size=10), 2)
q = subtraction(1.0, proportion)
half_numerator = product(proportion, q)
z = normalProbabilityDensity(denominator)
numerator = product(power(z, 2), half_numerator)
n = division(numerator, denominator)
return n
def sample_size_unknown(percent, confidence, width):
confidence_int = division(confidence, 2)
zscore = normalProbabilityDensity(confidence_int)
error = division(width, 2)
p = subtraction(1, percent)
pTimesq = product(percent, p)
zDivideError = division(zscore, error)
powerZdivError = power(zDivideError, 2)
return product(pTimesq, powerZdivError)
def sampleCorrelation(dataX, dataY):
#dataX= []
#dataY = []
meanX = mean(dataX)
meanY = mean(dataY)
deviationX = standard_deviation(dataX)
deviationY = standard_deviation(dataY)
rNumerator = 0.0
for i in range(len(dataX)):
rNumerator += product(subtraction(dataX[i], meanX),
subtraction(dataY[i], meanY))
rDenominator = product(deviationX, deviationY)
r = division(rNumerator, rDenominator)
return r
def margin_error(data, sample_size):
zscore = zScore(data)
standardDeviation = standard_deviation(data)
denominator = root(sample_size, 2)
willMultiply = division(standardDeviation, denominator)
marginOfError = product(zscore, willMultiply)
return marginOfError
def samplesizeKnownPop(data, confidence, error):
z = division(confidence, 2)
ztable_value = normalProbabilityDensity(z)
standardDeviation = standard_deviation(data)
ztableTimesStandard = product(ztable_value, standardDeviation)
nextStep = division(ztableTimesStandard, error)
sample_size = power(nextStep, 2)
return sample_size
def normalProbabilityDensity(x):
constant = division(1.0, math.sqrt(2 * math.pi))
return product(constant, math.exp((-x**2) / 2.0))