Get percentile points from a huge list - python

I have a huge list (45M+ data poitns), with numerical values:
[78,0,5,150,9000,5,......,25,9,78422...]
I can easily get the maximum and minimum values, the number of these values, and the sum of them:
file_handle=open('huge_data_file.txt','r')
sum_values=0
min_value=None
max_value=None
for i,line in enumerate(file_handle):
value=int(line[:-1])
if min_value==None or value<min_value:
min_value=value
if max_value==None or value>max_value:
max_value=value
sum_values+=value
average_value=float(sum_values)/i
However, this is not what I need. I need a list of 10 numbers, where the number of data points between each two consecutive points is equal, for example
median points [0,30,120,325,912,1570,2522,5002,7025,78422]
and we have the number of data points between 0 and 30 or between 30 and 120 to be almost 4.5 million data points.
How can we do this?
=============================
EDIT:
I am well aware that we will need to sort the data. The problem is that I cannot fit all this data in one variable in memory, but I need to read it sequentially from a generator (file_handle)

If you are happy with an approximation, here is a great (and fairly easy to implement) algorithm for computing quantiles from stream data: "Space-Efficient Online Computation of Quantile Summaries" by Greenwald and Khanna.

The silly numpy approach:
import numpy as np
# example data (produced by numpy but converted to a simple list)
datalist = list(np.random.randint(0, 10000000, 45000000))
# converted back to numpy array (start here with your data)
arr = np.array(datalist)
np.percentile(arr, 10), np.percentile(arr, 20), np.percentile(arr, 30)
# ref:
# http://docs.scipy.org/doc/numpy-dev/reference/generated/numpy.percentile.html
You can also hack something together where you just do like:
arr.sort()
# And then select the 10%, 20% etc value, add some check for equal amount of
# numbers within a bin and then calculate the average, excercise for reader :-)
The thing is that calling this function several times will slow it down, so really, just sort the array and then select the elements yourself.

As you said in the comments that you want a solution that can scale to larger datasets then can be stored in RAM, feed the data into an SQLlite3 database. Even if your data set is 10GB and you only have 8GB RAM a SQLlite3 database should still be able to sort the data and give it back to you in order.
The SQLlite3 database gives you a generator over your sorted data.
You might also want to look into going beyond python and take some other database solution.

Here's a pure-python implementation of the partitioned-on-disk sort. It's slow, ugly code, but it works and hopefully each stage is relatively clear (the merge stage is really ugly!).
#!/usr/bin/env python
import os
def get_next_int_from_file(f):
l = f.readline()
if not l:
return None
return int(l.strip())
MAX_SAMPLES_PER_PARTITION = 1000000
PARTITION_FILENAME = "_{}.txt"
# Partition data set
part_id = 0
eof = False
with open("data.txt", "r") as fin:
while not eof:
print "Creating partition {}".format(part_id)
with open(PARTITION_FILENAME.format(part_id), "w") as fout:
for _ in range(MAX_SAMPLES_PER_PARTITION):
line = fin.readline()
if not line:
eof = True
break
fout.write(line)
part_id += 1
num_partitions = part_id
# Sort each partition
for part_id in range(num_partitions):
print "Reading unsorted partition {}".format(part_id)
with open(PARTITION_FILENAME.format(part_id), "r") as fin:
samples = [int(line.strip()) for line in fin.readlines()]
print "Disk-Deleting unsorted {}".format(part_id)
os.remove(PARTITION_FILENAME.format(part_id))
print "In-memory sorting partition {}".format(part_id)
samples.sort()
print "Writing sorted partition {}".format(part_id)
with open(PARTITION_FILENAME.format(part_id), "w") as fout:
fout.writelines(["{}\n".format(sample) for sample in samples])
# Merge-sort the partitions
# NB This is a very inefficient implementation!
print "Merging sorted partitions"
part_files = []
part_next_int = []
num_lines_out = 0
# Setup data structures for the merge
for part_id in range(num_partitions):
fin = open(PARTITION_FILENAME.format(part_id), "r")
next_int = get_next_int_from_file(fin)
if next_int is None:
continue
part_files.append(fin)
part_next_int.append(next_int)
with open("data_sorted.txt", "w") as fout:
while part_files:
# Find the smallest number across all files
min_number = None
min_idx = None
for idx in range(len(part_files)):
if min_number is None or part_next_int[idx] < min_number:
min_number = part_next_int[idx]
min_idx = idx
# Now add that number, and move the relevent file along
fout.write("{}\n".format(min_number))
num_lines_out += 1
if num_lines_out % MAX_SAMPLES_PER_PARTITION == 0:
print "Merged samples: {}".format(num_lines_out)
next_int = get_next_int_from_file(part_files[min_idx])
if next_int is None:
# Remove this partition, it's now finished
del part_files[min_idx:min_idx + 1]
del part_next_int[min_idx:min_idx + 1]
else:
part_next_int[min_idx] = next_int
# Cleanup partition files
for part_id in range(num_partitions):
os.remove(PARTITION_FILENAME.format(part_id))

My code a proposal for finding the result without needing much space. In testing it found a quantile value in 7 minutes 51 seconds for a dataset of size 45 000 000.
from bisect import bisect_left
class data():
def __init__(self, values):
random.shuffle(values)
self.values = values
def __iter__(self):
for i in self.values:
yield i
def __len__(self):
return len(self.values)
def sortedValue(self, percentile):
val = list(self)
val.sort()
num = int(len(self)*percentile)
return val[num]
def init():
numbers = data([x for x in range(1,1000000)])
print(seekPercentile(numbers, 0.1))
print(numbers.sortedValue(0.1))
def seekPercentile(numbers, percentile):
lower, upper = minmax(numbers)
maximum = upper
approx = _approxPercentile(numbers, lower, upper, percentile)
return neighbor(approx, numbers, maximum)
def minmax(list):
minimum = float("inf")
maximum = float("-inf")
for num in list:
if num>maximum:
maximum = num
if num<minimum:
minimum = num
return minimum, maximum
def neighbor(approx, numbers, maximum):
dif = maximum
for num in numbers:
if abs(approx-num)<dif:
result = num
dif = abs(approx-num)
return result
def _approxPercentile(numbers, lower, upper, percentile):
middles = []
less = []
magicNumber = 10000
step = (upper - lower)/magicNumber
less = []
for i in range(1, magicNumber-1):
middles.append(lower + i * step)
less.append(0)
for num in numbers:
index = bisect_left(middles,num)
if index<len(less):
less[index]+= 1
summing = 0
for index, testVal in enumerate(middles):
summing += less[index]
if summing/len(numbers) < percentile:
print(" Change lower from "+str(lower)+" to "+ str(testVal))
lower = testVal
if summing/len(numbers) > percentile:
print(" Change upper from "+str(upper)+" to "+ str(testVal))
upper = testVal
break
precision = 0.01
if (lower+precision)>upper:
return lower
else:
return _approxPercentile(numbers, lower, upper, percentile)
init()
I edited my code a bit and I now think that this way works at least decently even when it's not optimal.

Related

Comparing Two Almost the Same Audio File(.MP3)

I have a lot of these small .mp3 files, and what I want to get here is to check if two audio speaks the same alphabet.
For example:
if audio_is_same("file1.mp3", "file2.mp3"):
print("Same")
else:
print("Different")
And here are some
Audio Samples (Some folders are empty.)
Since these audio are almost the same, I think it is possible to do it with a simple way?
Would training a audio recognition module simpler?
Audio files are just binary when you open them, so you can just compare the files after reading them in.
def compare_audio(file1, file2):
is_same = open(“file1”, "rb").read() == open(“file2”, "rb").read()
if is_same:
print('Same')
else:
print('Different')
If you have large files, compare in chunks as mentioned in the below link.
https://www.quora.com/How-do-I-compare-two-binary-files-in-Python
If you want to get some sort of similarity between the two you can use built in similarity function or some kind of model
from difflib import SequenceMatcher
threshold = 0.8
def similar(a, b):
return SequenceMatcher(None, a, b).ratio()
def compare_audio(file1, file2):
file1 = open(“file1”, "rb").read()
file2 = open(“file2”, "rb").read()
sim_ratio = similar(file1, file2)
if sim_ratio > threshold:
print('Same')
else:
print('Different')
You would need to decide what an appropriate threshold is.
I don't know specifically what differences are you looking for, but below there's a code to get a number from 0 to 100 for the similarity from two audio files using python, it works by generating fingerprints from audio files and comparing them based out of them using cross correlation
It requires Chromaprint and FFMPEG installed, also it doesn't work for short audio files, if this is a problem, you can always reduce the speed of the audio like in this guide, be aware this is going to add a little noise.
# correlation.py
import subprocess
import numpy
# seconds to sample audio file for
sample_time = 500# number of points to scan cross correlation over
span = 150# step size (in points) of cross correlation
step = 1# minimum number of points that must overlap in cross correlation
# exception is raised if this cannot be met
min_overlap = 20# report match when cross correlation has a peak exceeding threshold
threshold = 0.5
# calculate fingerprint
def calculate_fingerprints(filename):
fpcalc_out = subprocess.getoutput('fpcalc -raw -length %i %s' % (sample_time, filename))
fingerprint_index = fpcalc_out.find('FINGERPRINT=') + 12
# convert fingerprint to list of integers
fingerprints = list(map(int, fpcalc_out[fingerprint_index:].split(',')))
return fingerprints
# returns correlation between lists
def correlation(listx, listy):
if len(listx) == 0 or len(listy) == 0:
# Error checking in main program should prevent us from ever being
# able to get here.
raise Exception('Empty lists cannot be correlated.')
if len(listx) > len(listy):
listx = listx[:len(listy)]
elif len(listx) < len(listy):
listy = listy[:len(listx)]
covariance = 0
for i in range(len(listx)):
covariance += 32 - bin(listx[i] ^ listy[i]).count("1")
covariance = covariance / float(len(listx))
return covariance/32
# return cross correlation, with listy offset from listx
def cross_correlation(listx, listy, offset):
if offset > 0:
listx = listx[offset:]
listy = listy[:len(listx)]
elif offset < 0:
offset = -offset
listy = listy[offset:]
listx = listx[:len(listy)]
if min(len(listx), len(listy)) < min_overlap:
# Error checking in main program should prevent us from ever being
# able to get here.
return
#raise Exception('Overlap too small: %i' % min(len(listx), len(listy)))
return correlation(listx, listy)
# cross correlate listx and listy with offsets from -span to span
def compare(listx, listy, span, step):
if span > min(len(listx), len(listy)):
# Error checking in main program should prevent us from ever being
# able to get here.
raise Exception('span >= sample size: %i >= %i\n' % (span, min(len(listx), len(listy))) + 'Reduce span, reduce crop or increase sample_time.')
corr_xy = []
for offset in numpy.arange(-span, span + 1, step):
corr_xy.append(cross_correlation(listx, listy, offset))
return corr_xy
# return index of maximum value in list
def max_index(listx):
max_index = 0
max_value = listx[0]
for i, value in enumerate(listx):
if value > max_value:
max_value = value
max_index = i
return max_index
def get_max_corr(corr, source, target):
max_corr_index = max_index(corr)
max_corr_offset = -span + max_corr_index * step
print("max_corr_index = ", max_corr_index, "max_corr_offset = ", max_corr_offset)
# report matches
if corr[max_corr_index] > threshold:
print(('%s and %s match with correlation of %.4f at offset %i' % (source, target, corr[max_corr_index], max_corr_offset)))
def correlate(source, target):
fingerprint_source = calculate_fingerprints(source)
fingerprint_target = calculate_fingerprints(target)
corr = compare(fingerprint_source, fingerprint_target, span, step)
max_corr_offset = get_max_corr(corr, source, target)
if __name__ == "__main__":
correlate(SOURCE_FILE, TARGET_FILE)
Code converted into python 3 from: https://shivama205.medium.com/audio-signals-comparison-23e431ed2207

Most frequently overlapping range - Python3.x

I'm a beginner, trying to write code listing the most frequently overlapping ranges in a list of ranges.
So, input is various ranges (#1 through #7 in the example figure; https://prntscr.com/kj80xl) and I would like to find the most common range (in the example 3,000- 4,000 in 6 out of 7 - 86 %). Actually, I would like to find top 5 most frequent.
Not all ranges overlap. Ranges are always positive and given as integers with 1 distance (standard range).
What I have now is only code comparing one sequence to another and returning the overlap, but after that I'm stuck.
def range_overlap(range_x,range_y):
x = (range_x[0], (range_x[-1])+1)
y = (range_y[0], (range_y[-1])+1)
overlap = (max(x[0],y[0]),min(x[-1],(y[-1])))
if overlap[0] <= overlap[1]:
return range(overlap[0], overlap[1])
else:
return "Out of range"
I would be very grateful for any help.
Better solution
I came up with a simpler solution (at least IMHO) so here it is:
def get_abs_min(ranges):
return min([min(r) for r in ranges])
def get_abs_max(ranges):
return max([max(r) for r in ranges])
def count_appearances(i, ranges):
return sum([1 for r in ranges if i in r])
def create_histogram(ranges):
keys = [str(i) for i in range(len(ranges) + 1)]
histogram = dict.fromkeys(keys)
results = []
min = get_abs_min(range_list)
max = get_abs_max(range_list)
for i in range(min, max):
count = str(count_appearances(i, ranges))
if histogram[count] is None:
histogram[count] = dict(start=i, end=None)
elif histogram[count]['end'] is None:
histogram[count]['end'] = i
elif histogram[count]['end'] == i - 1:
histogram[count]['end'] = i
else:
start = histogram[count]['start']
end = histogram[count]['end']
results.append((range(start, end + 1), count))
histogram[count]['start'] = i
histogram[count]['end'] = None
for count, d in histogram.items():
if d is not None and d['start'] is not None and d['end'] is not None:
results.append((range(d['start'], d['end'] + 1), count))
return results
def main(ranges, top):
appearances = create_histogram(ranges)
return sorted(appearances, key=lambda t: t[1], reverse=True)[:top]
The idea here is as simple as iterating through a superposition of all the ranges and building a histogram of appearances (e.g. the number of original ranges this current i appears in)
After that just sort and slice according to the chosen size of the results.
Just call main with the ranges and the top number you want (or None if you want to see all results).
OLDER EDITS BELOW
I (almost) agree with #Kasramvd's answer.
here is my take on it:
from collections import Counter
from itertools import combinations
def range_overlap(x, y):
common_part = list(set(x) & set(y))
if common_part:
return range(common_part[0], common_part[-1] +1)
else:
return False
def get_most_common(range_list, top_frequent):
overlaps = Counter(range_overlap(i, j) for i, j in
combinations(list_of_ranges, 2))
return [(r, i) for (r, i) in overlaps.most_common(top_frequent) if r]
you need to input the range_list and the number of top_frequent you want.
EDIT
the previous answer solved this question for all 2's combinations over the range list.
This edit is tested against your input and results with the correct answer:
from collections import Counter
from itertools import combinations
def range_overlap(*args):
sets = [set(r) for r in args]
common_part = list(set(args[0]).intersection(*sets))
if common_part:
return range(common_part[0], common_part[-1] +1)
else:
return False
def get_all_possible_combinations(range_list):
all_combos = []
for i in range(2, len(range_list)):
all_combos.append(combinations(range_list, i))
all_combos = [list(combo) for combo in all_combos]
return all_combos
def get_most_common_for_combo(combo):
return list(filter(None, [range_overlap(*option) for option in combo]))
def get_most_common(range_list, top_frequent):
all_overlaps = []
combos = get_all_possible_combinations(range_list)
for combo in combos:
all_overlaps.extend(get_most_common_for_combo(combo))
return [r for (r, i) in Counter(all_overlaps).most_common(top_frequent) if r]
And to get the results just run get_most_common(range_list, top_frequent)
Tested on my machine (ubunut 16.04 with python 3.5.2) with your input range_list and top_frequent = 5 with the results:
[range(3000, 4000), range(2500, 4000), range(1500, 4000), range(3000, 6000), range(1, 4000)]
You can first change your function to return a valid range in both cases so that you can use it in a set of comparisons. Also, since Python's range objects are not already created iterables but smart objects that only get start, stop and step attributes of a range and create the range on-demand, you can do a little change on your function as well.
def range_overlap(range_x,range_y):
rng = range(max(range_x.start, range_y.start),
min(range_x.stop, range_y.stop)+1)
if rng.start < rng.stop:
return rng.start, rng.stop
Now, if you have a set of ranges and you want to compare all the pairs you can use itertools.combinations to get all the pairs and then using range_overlap and collections.Counter you can find the number of overlapped ranges.
from collections import Counter
from itertools import combinations
overlaps = Counter(range_overlap(i,j) for i, j in
combinations(list_of_ranges, 2))

How to reduce a collection of ranges to a minimal set of ranges [duplicate]

This question already has answers here:
Union of multiple ranges
(5 answers)
Closed 7 years ago.
I'm trying to remove overlapping values from a collection of ranges.
The ranges are represented by a string like this:
499-505 100-115 80-119 113-140 500-550
I want the above to be reduced to two ranges: 80-140 499-550. That covers all the values without overlap.
Currently I have the following code.
cr = "100-115 115-119 113-125 80-114 180-185 500-550 109-120 95-114 200-250".split(" ")
ar = []
br = []
for i in cr:
(left,right) = i.split("-")
ar.append(left);
br.append(right);
inc = 0
for f in br:
i = int(f)
vac = []
jnc = 0
for g in ar:
j = int(g)
if(i >= j):
vac.append(j)
del br[jnc]
jnc += jnc
print vac
inc += inc
I split the array by - and store the range limits in ar and br. I iterate over these limits pairwise and if the i is at least as great as the j, I want to delete the element. But the program doesn't work. I expect it to produce this result: 80-125 500-550 200-250 180-185
For a quick and short solution,
from operator import itemgetter
from itertools import groupby
cr = "499-505 100-115 80-119 113-140 500-550".split(" ")
fullNumbers = []
for i in cr:
a = int(i.split("-")[0])
b = int(i.split("-")[1])
fullNumbers+=range(a,b+1)
# Remove duplicates and sort it
fullNumbers = sorted(list(set(fullNumbers)))
# Taken From http://stackoverflow.com/questions/2154249
def convertToRanges(data):
result = []
for k, g in groupby(enumerate(data), lambda (i,x):i-x):
group = map(itemgetter(1), g)
result.append(str(group[0])+"-"+str(group[-1]))
return result
print convertToRanges(fullNumbers)
#Output: ['80-140', '499-550']
For the given set in your program, output is ['80-125', '180-185', '200-250', '500-550']
Main Possible drawback of this solution: This may not be scalable!
Let me offer another solution that doesn't take time linearly proportional to the sum of the range sizes. Its running time is linearly proportional to the number of ranges.
def reduce(range_text):
parts = range_text.split()
if parts == []:
return ''
ranges = [ tuple(map(int, part.split('-'))) for part in parts ]
ranges.sort()
new_ranges = []
left, right = ranges[0]
for range in ranges[1:]:
next_left, next_right = range
if right + 1 < next_left: # Is the next range to the right?
new_ranges.append((left, right)) # Close the current range.
left, right = range # Start a new range.
else:
right = max(right, next_right) # Extend the current range.
new_ranges.append((left, right)) # Close the last range.
return ' '.join([ '-'.join(map(str, range)) for range in new_ranges ]
This function works by sorting the ranges, then looking at them in order and merging consecutive ranges that intersect.
Examples:
print(reduce('499-505 100-115 80-119 113-140 500-550'))
# => 80-140 499-550
print(reduce('100-115 115-119 113-125 80-114 180-185 500-550 109-120 95-114 200-250'))
# => 80-125 180-185 200-250 500-550

Algorithm for matching objects

I have 1,000 objects, each object has 4 attribute lists: a list of words, images, audio files and video files.
I want to compare each object against:
a single object, Ox, from the 1,000.
every other object.
A comparison will be something like:
sum(words in common+ images in common+...).
I want an algorithm that will help me find the closest 5, say, objects to Ox and (a different?) algorithm to find the closest 5 pairs of objects
I've looked into cluster analysis and maximal matching and they don't seem to exactly fit this scenario. I don't want to use these method if something more apt exists, so does this look like a particular type of algorithm to anyone, or can anyone point me in the right direction to applying the algorithms I mentioned to this?
I made an example program for how to solve your first question. But you have to implement ho you want to compare images, audio and videos. And I assume every object has the same length for all lists. To answer your question number two it would be something similar, but with a double loop.
import numpy as np
from random import randint
class Thing:
def __init__(self, words, images, audios, videos):
self.words = words
self.images = images
self.audios = audios
self.videos = videos
def compare(self, other):
score = 0
# Assuming the attribute lists have the same length for both objects
# and that they are sorted in the same manner:
for i in range(len(self.words)):
if self.words[i] == other.words[i]:
score += 1
for i in range(len(self.images)):
if self.images[i] == other.images[i]:
score += 1
# And so one for audio and video. You have to make sure you know
# what method to use for determining when an image/audio/video are
# equal.
return score
N = 1000
things = []
words = np.random.randint(5, size=(N,5))
images = np.random.randint(5, size=(N,5))
audios = np.random.randint(5, size=(N,5))
videos = np.random.randint(5, size=(N,5))
# For testing purposes I assign each attribute to a list (array) containing
# five random integers. I don't know how you actually intend to do it.
for i in xrange(N):
things.append(Thing(words[i], images[i], audios[i], videos[i]))
# I will assume that object number 999 (i=999) is the Ox:
ox = 999
scores = np.zeros(N - 1)
for i in xrange(N - 1):
scores[i] = (things[ox].compare(things[i]))
best = np.argmax(scores)
print "The most similar thing is thing number %d." % best
print
print "Ox attributes:"
print things[ox].words
print things[ox].images
print things[ox].audios
print things[ox].videos
print
print "Best match attributes:"
print things[ox].words
print things[ox].images
print things[ox].audios
print things[ox].videos
EDIT:
Now here is the same program modified sligthly to answer your second question. It turned out to be very simple. I basically just needed to add 4 lines:
Changing scores into a (N,N) array instead of just (N).
Adding for j in xrange(N): and thus creating a double loop.
if i == j:
break
where 3. and 4. is just to make sure that I only compare each pair of things once and not twice and don't compary any things with themselves.
Then there is a few more lines of code that is needed to extract the indices of the 5 largest values in scores. I also reformated the printing so it will be easy to confirm by eye that the printed pairs are actually very similar.
Here comes the new code:
import numpy as np
class Thing:
def __init__(self, words, images, audios, videos):
self.words = words
self.images = images
self.audios = audios
self.videos = videos
def compare(self, other):
score = 0
# Assuming the attribute lists have the same length for both objects
# and that they are sorted in the same manner:
for i in range(len(self.words)):
if self.words[i] == other.words[i]:
score += 1
for i in range(len(self.images)):
if self.images[i] == other.images[i]:
score += 1
for i in range(len(self.audios)):
if self.audios[i] == other.audios[i]:
score += 1
for i in range(len(self.videos)):
if self.videos[i] == other.videos[i]:
score += 1
# You have to make sure you know what method to use for determining
# when an image/audio/video are equal.
return score
N = 1000
things = []
words = np.random.randint(5, size=(N,5))
images = np.random.randint(5, size=(N,5))
audios = np.random.randint(5, size=(N,5))
videos = np.random.randint(5, size=(N,5))
# For testing purposes I assign each attribute to a list (array) containing
# five random integers. I don't know how you actually intend to do it.
for i in xrange(N):
things.append(Thing(words[i], images[i], audios[i], videos[i]))
################################################################################
############################# This is the new part: ############################
################################################################################
scores = np.zeros((N, N))
# Scores will become a triangular matrix where scores[i, j]=value means that
# value is the number of attrributes thing[i] and thing[j] have in common.
for i in xrange(N):
for j in xrange(N):
if i == j:
break
# Break the loop here because:
# * When i==j we would compare thing[i] with itself, and we don't
# want that.
# * For every combination where j>i we would repeat all the
# comparisons for j<i and create duplicates. We don't want that.
scores[i, j] = (things[i].compare(things[j]))
# I want the 5 most similar pairs:
n = 5
# This list will contain a tuple for each of the n most similar pairs:
best_list = []
for k in xrange(n):
ij = np.argmax(scores) # Returns a single integer: ij = i*n + j
i = ij / N
j = ij % N
best_list.append((i, j))
# Erease this score so that on next iteration the second largest score
# is found:
scores[i, j] = 0
for k, (i, j) in enumerate(best_list):
# The number 1 most similar pair is the BEST match of all.
# The number N most similar pair is the WORST match of all.
print "The number %d most similar pair is thing number %d and %d." \
% (k+1, i, j)
print "Thing%4d:" % i, \
things[i].words, things[i].images, things[i].audios, things[i].videos
print "Thing%4d:" % j, \
things[j].words, things[j].images, things[j].audios, things[j].videos
print
If your comparison works with "create a sum of all features and find those which the closest sum", there is a simple trick to get close objects:
Put all objects into an array
Calculate all the sums
Sort the array by sum.
If you take any index, the objects close to it will now have a close index as well. So to find the 5 closest objects, you just need to look at index+5 to index-5 in the sorted array.

Sort (and sorted) not sorting

I have a data file that's built the following way:
source_id, target_id, impressions, clicks
on which I add the following columns:
pair - a tuple of the source and target
CTR - basically clicks/impressions
Lower Bound
Upper Bound
Lower/Upper bound are calculated values (it's irrelevant to my question, but for the curious ones these are the bounds for the Wilson confidence interval.
The thing is, I'm trying to sort the list by the lower bound (position = 6), descending. Tried several things (sort/sorted, using lambda vs. using itemgetter, creating a new list w/o the header and try to sort just that) and still it appears nothing changes. I have the code below.
import csv
from math import sqrt
from operator import itemgetter
#----- Read CSV ----------------------------------------------------------------
raw_data_csv = open('rawdile', "rb")
raw_reader = csv.reader(raw_data_csv)
# transform the values to ints.
raw_data = []
for rownum,row in enumerate(list(raw_reader)):
if rownum == 0: # Header
raw_data.append(row)
else:
r = [] # Col header
r.extend([int(x) for x in row]) # Transforming the values to ints
raw_data.append(r)
# Add cols for pairs (as tuple) and CTR
raw_data[0].append("pair")
for row in raw_data[1:]:
row.append((row[0],row[1])) # tuple
# row.append(float(row[3])/row[2]) # CTR
# ------------------------------------------------------------------------------
z = 1.95996398454005
def confidence(n, clicks):
if n == 0:
return 0
phat = float(clicks) / n
l_bound = ((phat + z*z/(2*n) - z * sqrt((phat*(1-phat)+z*z/(4*n))/n))/(1+z*z/n)) # lower bound
u_bound = ((phat + z*z/(2*n) + z * sqrt((phat*(1-phat)+z*z/(4*n))/n))/(1+z*z/n)) # upper bound
return phat, l_bound, u_bound
raw_data[0].extend(["CTR","Lower Bound","Upper Bound"])
for row in raw_data[1:]:
phat, l_bound, u_bound = confidence(row[2],row[3])
row.extend([phat, l_bound, u_bound])
# raw_data[1:].sort(key=lambda x: x[6], reverse=True)
sorted(raw_data[1:], key=itemgetter(6), reverse=True)
outputfile= open('outputfile.csv', 'wb')
wr = csv.writer(outputfile,quoting = csv.QUOTE_ALL)
wr.writerows(raw_data)
raw_data_csv.close()
outputfile.close()
Can anybody tell why?
Thanks!
You are sorting a slice in one attempt (which creates a new list object), and in your other attempt you ignore the return value of sorted().
You cannot sort part of a list like that; create a new list by concatenating instead:
rows = rows[:1] + sorted(raw_data[1:], key=itemgetter(6), reverse=True)

Categories