如何在Python中实现R的p.adjust

Question

我有一个p值列表,我想计算fDR的多重比较的调整p值.在R中,我可以使用:

pval <- read.csv("my_file.txt",header=F,sep="\t")
pval <- pval[,1]
FDR <- p.adjust(pval, method= "BH")
print(length(pval[FDR<0.1]))
write.table(cbind(pval, FDR),"pval_FDR.txt",row.names=F,sep="\t",quote=F )

如何在Python中实现此代码？以下是我在Google的帮助下在Python中的可行尝试:

pvalue_list [2.26717873145e-10, 1.36209234286e-11 , 0.684342083821...] # my pvalues
pvalue_lst = [v.r['p.value'] for v in pvalue_list]
p_adjust = R.r['p.adjust'](R.FloatVector(pvalue_lst),method='BH')
for v in p_adjust:
    print v

上面的代码抛出了一个AttributeError: 'float' object has no attribute 'r'错误.任何人都可以帮助指出我的问题吗？在此先感谢您的帮助!

Answer 1

如果您希望确定从R获得的内容,您还可以表明您希望使用R包'stats'中的功能:

from rpy2.robjects.packages import importr
from rpy2.robjects.vectors import FloatVector

stats = importr('stats')

p_adjust = stats.p_adjust(FloatVector(pvalue_list), method = 'BH')

Answer 2

这个问题有点陈旧,但在Python的statsmodels中有多种比较修正.我们有

http://statsmodels.sourceforge.net/devel/generated/statsmodels.sandbox.stats.multicomp.multipletests.html#statsmodels.sandbox.stats.multicomp.multipletests

Answer 3

这是我使用的内部功能:

def correct_pvalues_for_multiple_testing(pvalues, correction_type = "Benjamini-Hochberg"):                
    """                                                                                                   
    consistent with R - print correct_pvalues_for_multiple_testing([0.0, 0.01, 0.029, 0.03, 0.031, 0.05, 0.069, 0.07, 0.071, 0.09, 0.1]) 
    """
    from numpy import array, empty                                                                        
    pvalues = array(pvalues) 
    n = float(pvalues.shape[0])                                                                           
    new_pvalues = empty(n)
    if correction_type == "Bonferroni":                                                                   
        new_pvalues = n * pvalues
    elif correction_type == "Bonferroni-Holm":                                                            
        values = [ (pvalue, i) for i, pvalue in enumerate(pvalues) ]                                      
        values.sort()
        for rank, vals in enumerate(values):                                                              
            pvalue, i = vals
            new_pvalues[i] = (n-rank) * pvalue                                                            
    elif correction_type == "Benjamini-Hochberg":                                                         
        values = [ (pvalue, i) for i, pvalue in enumerate(pvalues) ]                                      
        values.sort()
        values.reverse()                                                                                  
        new_values = []
        for i, vals in enumerate(values):                                                                 
            rank = n - i
            pvalue, index = vals                                                                          
            new_values.append((n/rank) * pvalue)                                                          
        for i in xrange(0, int(n)-1):  
            if new_values[i] < new_values[i+1]:                                                           
                new_values[i+1] = new_values[i]                                                           
        for i, vals in enumerate(values):
            pvalue, index = vals
            new_pvalues[index] = new_values[i]                                                                                                                  
    return new_pvalues

Answer 4

使用Python的numpy库,根本没有调用R,这是BH方法的一个相当有效的实现:

import numpy as np

def p_adjust_bh(p):
    """Benjamini-Hochberg p-value correction for multiple hypothesis testing."""
    p = np.asfarray(p)
    by_descend = p.argsort()[::-1]
    by_orig = by_descend.argsort()
    steps = float(len(p)) / np.arange(len(p), 0, -1)
    q = np.minimum(1, np.minimum.accumulate(steps * p[by_descend]))
    return q[by_orig]

(基于R代码BondedDust发布)