How to implement R's p.adjust in Python

Question

I have a list of p-values and I would like to calculate the adjust p-values for multiple comparisons for the FDR. In R, I can use:

pval <- read.csv(\"my_f         


        

Answer 1

Old question, but here's a translation of the R FDR code in python (which is probably fairly inefficient):

def FDR(x):
    """
    Assumes a list or numpy array x which contains p-values for multiple tests
    Copied from p.adjust function from R  
    """
    o = [i[0] for i in sorted(enumerate(x), key=lambda v:v[1],reverse=True)]
    ro = [i[0] for i in sorted(enumerate(o), key=lambda v:v[1])]
    q = sum([1.0/i for i in xrange(1,len(x)+1)])
    l = [q*len(x)/i*x[j] for i,j in zip(reversed(xrange(1,len(x)+1)),o)]
    l = [l[k] if l[k] < 1.0 else 1.0 for k in ro]
    return l