def __init__(self, x, y, nocopy=None):
if len(x) != len(y): raise ValueError, 'x, y length mismatch!'
if len(x) < 2: raise ValueError, 'Need at least n=2!'
# Handle copying first so the datstt call doesn't duplicate
# any needed conversion of a sequence to an array.
if nocopy:
self.x = x
self.y = y
else:
try:
self.x = x.copy()
self.y = y.copy()
except AttributeError:
self.x, self.y = array(x), array(y)
self.xbar, self.xstd, self.ybar, self.ystd, \
self.covar, self.rho, error = _regress.datstt(self.x, self.y)
if error == 0:
raise ValueError, 'Input data has zero Sum(x-xbar)(y-ybar)!'
self.xstd = sqrt(self.xstd)
self.ystd = sqrt(self.ystd)
def __init__(self, x, y, nocopy=None):
if len(x) != len(y): raise ValueError, 'x, y length mismatch!'
if len(x) < 2: raise ValueError, 'Need at least n=2!'
# Handle copying first so the datstt call doesn't duplicate
# any needed conversion of a sequence to an array.
if nocopy:
self.x = x
self.y = y
else:
try:
self.x = x.copy()
self.y = y.copy()
except AttributeError:
self.x, self.y = array(x), array(y)
self.xbar, self.xstd, self.ybar, self.ystd, \
self.covar, self.rho, error = _regress.datstt(self.x, self.y)
if error == 0:
raise ValueError, 'Input data has zero Sum(x-xbar)(y-ybar)!'
self.xstd = sqrt(self.xstd)
self.ystd = sqrt(self.ystd)
def bistats(x, y):
"""
Basic statistics summarizing a bivariate distribution.
xbar, xstd, ybar, ystd, covar, rho = bistats(x,y)
xbar, ybar are the sample means
xstd, ystd are the sample standard deviations (with 1/n
rather than 1/(n-1) factors in the variance sum)
covar is the covariance
rho is Pearson's linear correlation coefficient
"""
if len(x) != len(y): raise ValueError, 'x, y length mismatch!'
if len(x) < 2: raise ValueError, 'Need at least n=2!'
xbar, xstd, ybar, ystd, covar, rho, ierr = _regress.datstt(x, y)
if ierr == 0:
raise ValueError, 'Input data has zero Sum(x-xbar)(y-ybar)!'
xstd = sqrt(xstd)
ystd = sqrt(ystd)
return xbar, xstd, ybar, ystd, covar, rho
def bistats(x, y):
"""
Basic statistics summarizing a bivariate distribution.
xbar, xstd, ybar, ystd, covar, rho = bistats(x,y)
xbar, ybar are the sample means
xstd, ystd are the sample standard deviations (with 1/n
rather than 1/(n-1) factors in the variance sum)
covar is the covariance
rho is Pearson's linear correlation coefficient
"""
if len(x) != len(y): raise ValueError, 'x, y length mismatch!'
if len(x) < 2: raise ValueError, 'Need at least n=2!'
xbar, xstd, ybar, ystd, covar, rho, ierr = _regress.datstt(x,y)
if ierr == 0:
raise ValueError, 'Input data has zero Sum(x-xbar)(y-ybar)!'
xstd = sqrt(xstd)
ystd = sqrt(ystd)
return xbar, xstd, ybar, ystd, covar, rho
