calculate Pearson correlation along with the confidence interval using scipy and numpy

pearsonr_ci.py

import numpy as np
from scipy import stats

def pearsonr_ci(x,y,alpha=0.05):
    ''' calculate Pearson correlation along with the confidence interval using scipy and numpy

    Parameters
    ----------
    x, y : iterable object such as a list or np.array
      Input for correlation calculation
    alpha : float
      Significance level. 0.05 by default

    Returns
    -------
    r : float
      Pearson's correlation coefficient
    pval : float
      The corresponding p value
    lo, hi : float
      The lower and upper bound of confidence intervals
    '''

    r, p = stats.pearsonr(x,y)
    r_z = np.arctanh(r)
    se = 1/np.sqrt(x.size-3)
    z = stats.norm.ppf(1-alpha/2)
    lo_z, hi_z = r_z-z*se, r_z+z*se
    lo, hi = np.tanh((lo_z, hi_z))
    return r, p, lo, hi