Converting R code to python

[mcva]

I´m trying to convert the following R code to python but theres something wrong with my code results are not the same can you take a look at it? Thank you.

With this input vector or list:
wtr = c(1,1,10,12,11,1) or wtr = [1,1,10,12,11,1]

Output:
The correct ouput should be 4 in R and 3 in python.

The output is an index of a value therefore there´s a difference in the output obtained with each code.



R code:
# Finds the local peaks in a vector. Checks the optionally supplied threshold
# for minimum height.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

findPeaks <- function(dataIn, thresh=0){           varL = length(dataIn);     locs = rep(FALSE, varL);     peaks= rep(NaN, varL);           for(i in 2:varL-1){         pkI = which.max(dataIn[(i-1):(i+1)])         posPeak = max(dataIn[(i-1):(i+1)]);                   if(pkI == 2) {             peaks[i] = posPeak;             locs[i]  = TRUE;         }     }           inds = 1:varL;     locs = inds[locs];     peaks= peaks[locs];           # remove all below threshold value           useI = peaks > thresh;     locs = locs[useI];           return(locs) }

python code:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

def findPeaks(dataIn, thresh=0.0):           varL = len(dataIn)     locs = [False]*varL     peaks = [None]*varL     for i in range(1,varL):         pKi = np.argmax(dataIn[(i-1):(i+1)])         posPeak = np.max(dataIn[(i-1):(i+1)])         if(pKi == 1):             peaks[i] = posPeak             locs[i] = True       locs = [i for i, e in enumerate(locs) if e == True]     peaks = [dataIn[i] for i in locs]           # remove all below threshold value     useI = [a for a in peaks if a > thresh]     locs = [i for i in range(varL) if dataIn[i] in useI]              return locs

[ichabod801]

The problem is lines 7 and 8 in the Python program. The '+1's should be '+2's to account for differences in R and Python slices. That is, findPeaks is looking for numbers that are at least as high as the two numbers on either side, so you need to be looking at a three number range.

That code was a bit clunky, so I redid it and removed the np dependency:

1 2 3 4 5 6 7

def find_peaks(data_in, thresh = 0.0):     peaks = [thresh - 1] * len(data_in)     for center in range(1, len(data_in)):         trip = data_in[(center - 1):(center + 2)]         if max(trip) == trip[1]:             peaks[center] = trip[1]     return [index for index, peak in enumerate(peaks) if peak > thresh]

I might be able to clean up that for loop more, let me poke around in itertools.

Edit: It cleans up the loop at the expense of the conditional.

1 2 3 4 5 6 7 8

def find_peaks(data_in, thresh = 0.0):     peaks = [thresh - 1]     for trip in zip(*[data_in[n:] for n in range(3)]):         if max(trip) == trip[1]:             peaks.append(trip[1])         else:             peaks.append(thresh - 1)     return [index for index, peak in enumerate(peaks) if peak > thresh]

[mcva]

You are right,
Thank you so much,