CODEWITHSUNDEEP
pythonlistlist-comprehension
chase
chase

Reputation: 3782

simple python list comprehension

I have a feeling that there is a simple answer to this question, but I am having trouble getting the answer without having to write many many lines of code. I was hoping someone may give me something that will make my life easier in regards to list comprehensions in python.

What I have is this:

highest_peak = sorted([(self.Y[x]-self.background_model[x]) for x in peaks_index_list])[0]

How do I obtain the index (x) of the highest_peak from the peaks_index_list where this occurs?

index_of_highest_peak = sorted(???)[0]
index_of_second_highest_peak = sorted(???)[1]

Thanks for any help.

Upvotes: 1

Views: 171

Answers (4)

tdelaney
tdelaney

Reputation: 77407

(note: edited based on comments)

Add the index to the objects being sorted:

highest_peak, index = sorted([(self.Y[x]-self.background_model[x], x)
    for x in peaks_index_list])[0]

(originally I suggested using enumerate(), which is great when you want the index of the elements you are enumerating. In this case, the peaks_index_list is already the index the poster was after.)

highest_peak, index = sorted([(self.Y[x]-self.background_model[x], i)
    for i, x in enumerate(peaks_index_list]))[0]

Upvotes: 2

kojiro
kojiro

Reputation: 77187

I take it the peak height is self.Y[i] - self.background_model[i] for a peak index i? If all you want is the index of the highest peak:

max(peaks_index_list, key=lambda i: self.Y[i] - self.background_model[i])

Likewise, if you really want the sorted list, use the above key to sort peaks_index_list.

Upvotes: 3

danodonovan
danodonovan

Reputation: 20373

If you use numpy you can 

import numpy as np
highest_peak = np.asarray(highest_peak)

max_index = highest_peak.argmax()

Upvotes: 3

mgilson
mgilson

Reputation: 310287

You should be sorting a list of tuples of the form (value,index) --

info = sorted((self.Y[x]-self.background_model[x],x) for x in peaks_index_list)

Now your index of the highest peak is:

info[-1][1]

Or you can unpack it a little more nicely:

highest_peak_value,highest_peak_index = info[-1]

(highest peak is last in the sorted list since we didn't specify reverse=True)

The second highest peak index is:

info[-2][1]

and so on.

Upvotes: 2

Related Questions