Convert a R code into Python script

Question

I got the following R code and I need to convert it to python and run it in python environment, basically I have done this with rpy2 module, but it looks kind of dull with python doing the same things, so could someone find a better way to rewrite the following R code to an equivalent python script with the rpy2 module?

mymad <- function (x) 
{
    center <- median(x)
    y <- abs(x - center)
    n <- length(y)
    if (n == 0) 
        return(NA)
    half <- (n + 1)/2
    1.4826 * if (n%%2 == 1) {
        sort(y, partial = half)[half]
    }
    else {
        sum(sort(y, partial = c(half, half + 1))[c(half, half + 
            1)])/2
    }
}

Wok · Accepted Answer · 2010-08-06 07:56:57Z

7

You could have stated the purpose of your function, which is Median Absolute Deviation. What you call mymad is an approximation of the standard deviation of the population, based on the assumption of large samples of normally distributed variables.

According to this website:

def median(pool):
    copy = sorted(pool)
    size = len(copy)
    if size % 2 == 1:
        return copy[(size - 1) / 2]
    else:
        return (copy[size/2 - 1] + copy[size/2]) / 2

So, you want a function mad which would verify :

mad(x) == median(abs(x-median(x)))

Thanks to Elenaher (give his comment credits), here is the code:

def mad(x):
    return median([abs(val-median(x)) for val in x])

And then, I believe your are computing:

def mymad(x):
    return 1.4826*mad(x)

edited Aug 6, 2010 at 7:56

answered Aug 5, 2010 at 16:11

Wok

5,4038 gold badges47 silver badges73 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

ThR37 Over a year ago

The widely used package numpy provides the median function (numpy.median) so don't waste time in reinventing the wheel !

Mark Over a year ago

Am i missing something? Assuming x is a list of numbers, (x - median(x)), Python won't do vectorized math.

ThR37 Over a year ago

@Mark yes but numpy does it ! If x is a numpy array you can write x-np.median(x). Otherwise you can use list comprehension : median([abs(val-median(x)) for val in x])

ligwin Over a year ago

thanks, btw, is there functions called Maclist and MLEintvl in R? could not find it...

Wok Over a year ago

If you are satisfied with the answer, then accept it. And if you have other questions, then the best way is to ask them using the usual form and the right tags.

lgautier · Accepted Answer · 2010-08-06 20:34:48Z

3

Probably a little slower than a numpy/Python written one, but certainly faster to implement (as no wheel gets reinvented):

# requires rpy2 >= 2.1
from rpy2.robjects.packages import importr
stats = importr('stats')

stats.mad(x)

answered Aug 6, 2010 at 20:34

lgautier

11.6k31 silver badges43 bronze badges

Comments

user3433265 · Accepted Answer · 2014-03-18 13:23:45Z

2

import numpy
# x is the input array
x = numpy.array( [1,2,4,3,1,6,7,5,4,6,7], float ) }
# mad = median( | x - median(x) | )
mad =  numpy.median( numpy.abs( ( x - numpy.median( x ) ) )

answered Mar 18, 2014 at 13:23

user3433265

211 bronze badge

1 Comment

Parixit Over a year ago

add some description to answer.

Collectives™ on Stack Overflow

Convert a R code into Python script

3 Answers 3

5 Comments

Comments

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

5 Comments

Comments

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related