Duplicate each member in an iterator - python

Given an iterator i, I want an iterator that yields each element n times, i.e., the equivalent of this function
def duplicate(i, n):
for x in i:
for k in range(n):
yield x
Is there an one-liner for this?
Related question: duplicate each member in a list - python, but the zip solution doesn't work here.

This is my simple solution, if you want to duplicate each element same times. It returns a generator expression, which should be memory efficient.
def duplicate(i, n):
return (k for k in i for j in range(n))
An example usage could be,
print (list(duplicate(range(1, 10), 3)))
Which prints,
[1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6, 6, 7, 7, 7, 8, 8,
8, 9, 9, 9]

itertools.chain.from_iterable(itertools.izip(*itertools.tee(source, n)))
Example:
>>> x = (a**2 for a in xrange(5))
>>> list(itertools.chain.from_iterable(itertools.izip(*itertools.tee(x, 3))))
[0, 0, 0, 1, 1, 1, 4, 4, 4, 9, 9, 9, 16, 16, 16]
Another way:
itertools.chain.from_iterable(itertools.repeat(item, n) for item in source)
>>> x = (a**2 for a in xrange(5))
>>> list(itertools.chain.from_iterable(itertools.repeat(item, 3) for item in x))
[0, 0, 0, 1, 1, 1, 4, 4, 4, 9, 9, 9, 16, 16, 16]

Use a generator expression:
>>> x = (n for n in range(4))
>>> i = (v for v in x for _ in range(3))
>>> list(i)
[0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3]

Related

Get list of numbers then return the numbers in the list that are increasing [duplicate]

The aim is to find groups of increasing/monotonic numbers given a list of integers. Each item in the resulting group must be of a +1 increment from the previous item
Given an input:
x = [7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
I need to find groups of increasing numbers and achieve:
increasing_numbers = [(7,8,9,10), (0,1,2,3,4,5)]
And eventually also the number of increasing numbers:
len(list(chain(*increasing_numbers)))
And also the len of the groups:
increasing_num_groups_length = [len(i) for i in increasing_numbers]
I have tried the following to get the number of increasing numbers:
>>> from itertools import tee, chain
>>> def pairwise(iterable):
... a, b = tee(iterable)
... next(b, None)
... return zip(a, b)
...
>>> x = [8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6]
>>> set(list(chain(*[(i,j) for i,j in pairwise(x) if j-1==i])))
set([1, 2, 3, 4, 5, 6, 8, 9, 10, 11])
>>> len(set(list(chain(*[(i,j) for i,j in pairwise(x) if j-1==i]))))
10
But I'm unable to keep the order and the groups of increasing numbers.
How can I achieve the increasing_numbers groups of integer tuples and also the increasing_num_groups_length?
Also, is there a name for such/similar problem?
EDITED
I've came up with this solution but it's super verbose and I'm sure there's an easier way to achieve the increasing_numbers output:
>>> from itertools import tee, chain
>>> def pairwise(iterable):
... a, b = tee(iterable)
... next(b, None)
... return zip(a, b)
...
>>> x = [8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6]
>>> boundary = iter([0] + [i+1 for i, (j,k) in enumerate(pairwise(x)) if j+1!=k] + [len(x)])
>>> [tuple(x[i:next(boundary)]) for i in boundary]
[(8, 9, 10, 11), (1, 2, 3, 4, 5, 6)]
Is there a more pythonic / less verbose way to do this?
Another input/output example:
[in]:
[17, 17, 19, 20, 21, 22, 0, 1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13,
14, 14, 14, 28, 29, 30, 31, 32, 33, 34, 35, 36, 40]
[out]:
[(19, 20, 21, 22), (0, 1, 2), (4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14),
(28, 29, 30, 31, 32, 33, 34, 35, 36)]
EDIT:
Here's a code-golf solution (142 characters):
def f(x):s=[0]+[i for i in range(1,len(x)) if x[i]!=x[i-1]+1]+[len(x)];return [x[j:k] for j,k in [s[i:i+2] for i in range(len(s)-1)] if k-j>1]
Expanded version:
def igroups(x):
s = [0] + [i for i in range(1, len(x)) if x[i] != x[i-1] + 1] + [len(x)]
return [x[j:k] for j, k in [s[i:i+2] for i in range(len(s)-1)] if k - j > 1]
Commented version:
def igroups(x):
# find the boundaries where numbers are not consecutive
boundaries = [i for i in range(1, len(x)) if x[i] != x[i-1] + 1]
# add the start and end boundaries
boundaries = [0] + boundaries + [len(x)]
# take the boundaries as pairwise slices
slices = [boundaries[i:i + 2] for i in range(len(boundaries) - 1)]
# extract all sequences with length greater than one
return [x[start:end] for start, end in slices if end - start > 1]
Original solution:
Not sure whether this counts as "pythonic" or "not too verbose":
def igroups(iterable):
items = iter(iterable)
a, b = None, next(items, None)
result = [b]
while b is not None:
a, b = b, next(items, None)
if b is not None and a + 1 == b:
result.append(b)
else:
if len(result) > 1:
yield tuple(result)
result = [b]
print(list(igroups([])))
print(list(igroups([0, 0, 0])))
print(list(igroups([7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5])))
print(list(igroups([8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6])))
print(list(igroups([9, 1, 2, 3, 1, 1, 2, 3, 5])))
Output:
[]
[]
[(7, 8, 9, 10), (0, 1, 2, 3, 4, 5)]
[(8, 9, 10, 11), (1, 2, 3, 4, 5, 6)]
[(1, 2, 3), (1, 2, 3)]
A couple of different ways using itertools and numpy:
from itertools import groupby, tee, cycle
x = [17, 17, 19, 20, 21, 22, 0, 1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 14, 14, 28, 29, 30, 31, 32, 33, 34, 35,
36, 1, 2, 3, 4,34,54]
def sequences(l):
x2 = cycle(l)
next(x2)
grps = groupby(l, key=lambda j: j + 1 == next(x2))
for k, v in grps:
if k:
yield tuple(v) + (next((next(grps)[1])),)
print(list(sequences(x)))
[(19, 20, 21, 22), (0, 1, 2), (4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14), (28, 29, 30, 31, 32, 33, 34, 35, 36), (1, 2, 3, 4)]
Or using python3 and yield from:
def sequences(l):
x2 = cycle(l)
next(x2)
grps = groupby(l, key=lambda j: j + 1 == next(x2))
yield from (tuple(v) + (next((next(grps)[1])),) for k,v in grps if k)
print(list(sequences(x)))
Using a variation of my answer here with numpy.split :
out = [tuple(arr) for arr in np.split(x, np.where(np.diff(x) != 1)[0] + 1) if arr.size > 1]
print(out)
[(19, 20, 21, 22), (0, 1, 2), (4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14), (28, 29, 30, 31, 32, 33, 34, 35, 36), (1, 2, 3, 4)]
And similar to ekhumoro's answer:
def sequences(x):
it = iter(x)
prev, temp = next(it), []
while prev is not None:
start = next(it, None)
if prev + 1 == start:
temp.append(prev)
elif temp:
yield tuple(temp + [prev])
temp = []
prev = start
To get the length and the tuple:
def sequences(l):
x2 = cycle(l)
next(x2)
grps = groupby(l, key=lambda j: j + 1 == next(x2))
for k, v in grps:
if k:
t = tuple(v) + (next(next(grps)[1]),)
yield t, len(t)
def sequences(l):
x2 = cycle(l)
next(x2)
grps = groupby(l, lambda j: j + 1 == next(x2))
yield from ((t, len(t)) for t in (tuple(v) + (next(next(grps)[1]),)
for k, v in grps if k))
def sequences(x):
it = iter(x)
prev, temp = next(it), []
while prev is not None:
start = next(it, None)
if prev + 1 == start:
temp.append(prev)
elif temp:
yield tuple(temp + [prev]), len(temp) + 1
temp = []
prev = start
Output will be the same for all three:
[((19, 20, 21, 22), 4), ((0, 1, 2), 3), ((4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14), 11)
, ((28, 29, 30, 31, 32, 33, 34, 35, 36), 9), ((1, 2, 3, 4), 4)]
I think the most maintainable solution would be to make it simple:
def group_by(l):
res = [[l[0]]]
for i in range(1, len(l)):
if l[i-1] < l[i]:
res[-1].append(l[i])
else:
res.append([l[i]])
return res
This solution does not filter out single element sequences, but it can be easily implemented. Additionally, this has O(n) complexity. And you can make it an generator as well if you want.
By maintainable I mean code that is not an one-liner of 300 characters, with some convoluted expressions. Then maybe you would want to use Perl :). At least you will how the function behaves one year later.
>>> x = [7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
>>> print(group_by(x))
[[7, 8, 9, 10], [6], [0, 1, 2, 3, 4, 5]]
If two consecutive numbers are increasing by one I form a list (group) of tuples of those numbers.
When non-increasing and if the list (group) is non-empty, I unpack it and zip again to rebuild the pair of sequence which were broken by the zip. I use set comprehension to eliminate duplicate numbers.
def extract_increasing_groups(seq):
seq = tuple(seq)
def is_increasing(a,b):
return a + 1 == b
def unzip(seq):
return tuple(sorted({ y for x in zip(*seq) for y in x}))
group = []
for a,b in zip(seq[:-1],seq[1:]):
if is_increasing(a,b):
group.append((a,b))
elif group:
yield unzip(group)
group = []
if group:
yield unzip(group)
if __name__ == '__main__':
x = [17, 17, 19, 20, 21, 22, 0, 1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 14, 14, 28, 29, 30, 31, 32, 33, 34, 35, 36, 40]
for group in extract_increasing_groups(x):
print(group)
Simpler one using set;
from collections import namedtuple
from itertools import islice, tee
def extract_increasing_groups(iterable):
iter1, iter2 = tee(iterable)
iter2 = islice(iter2,1,None)
is_increasing = lambda a,b: a + 1 == b
Igroup = namedtuple('Igroup','group, len')
group = set()
for pair in zip(iter1, iter2):
if is_increasing(*pair):
group.update(pair)
elif group:
yield Igroup(tuple(sorted(group)),len(group))
group = set()
if group:
yield Igroup(tuple(sorted(group)), len(group))
if __name__ == '__main__':
x = [17, 17, 19, 20, 21, 22, 0, 1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 14, 14, 28, 29, 30, 31, 32, 33, 34, 35, 36, 40]
total = 0
for group in extract_increasing_groups(x):
total += group.len
print('Group: {}\nLength: {}'.format(group.group, group.len))
print('Total: {}'.format(total))
def igroups(L):
R=[[]]
[R[-1].append(L[i]) for i in range(len(L)) if (L[i-1]+1==L[i] if L[i-1]+1==L[i] else R.append([L[i]]))]
return [P for P in R if len(P)>1]
tests=[[],
[0, 0, 0],
[7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5],
[8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6],
[9, 1, 2, 3, 1, 1, 2, 3, 5],
[4,3,2,1,1,2,3,3,4,3],
[1, 4, 3],
[1],
[1,2],
[2,1]
]
for L in tests:
print(L)
print(igroups(L))
print("-"*10)
outputting the following:
[]
[]
----------
[0, 0, 0]
[]
----------
[7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
[[7, 8, 9, 10], [0, 1, 2, 3, 4, 5]]
----------
[8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6]
[[8, 9, 10, 11], [1, 2, 3, 4, 5, 6]]
----------
[9, 1, 2, 3, 1, 1, 2, 3, 5]
[[1, 2, 3], [1, 2, 3]]
----------
[4, 3, 2, 1, 1, 2, 3, 3, 4, 3]
[[1, 2, 3], [3, 4]]
----------
[1, 4, 3]
[]
----------
[1]
[]
----------
[1, 2]
[[1, 2]]
----------
[2, 1]
[]
----------
EDIT
My first attemp using itertools.groupby was a fail, sorry for that.
With itertools.groupby, the problem of partionning a list of integers L in sublists of adjacent and increasing consecutive items from L can be done with a one-liner. Nevertheless I don't know how pythonic it can be considered ;)
Here is the code with some simple tests:
[EDIT : now subsequences are increasing by 1, I missed this point the first time.]
from itertools import groupby
def f(i):
return L[i-1]+1==L[i]
def igroups(L):
return [[L[I[0]-1]]+[L[i] for i in I] for I in [I for (v,I) in [(k,[i for i in list(g)]) for (k, g) in groupby(range(1, len(L)), f)] if v]]
outputting:
tests=[
[0, 0, 0, 0],
[7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5],
[8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6],
[9, 1, 2, 3, 1, 1, 2, 3, 5],
[4,3,2,1,1,2,3,3,4,3],
[1, 4, 3],
[1],
[1,2, 2],
[2,1],
[0, 0, 0, 0, 2, 5, 5, 8],
]
for L in tests:
print(L)
print(igroups(L))
print('-'*10)
[0, 0, 0, 0]
[]
----------
[7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
[[7, 8, 9, 10], [0, 1, 2, 3, 4, 5]]
----------
[8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6]
[[8, 9, 10, 11], [1, 2, 3, 4, 5, 6]]
----------
[9, 1, 2, 3, 1, 1, 2, 3, 5]
[[1, 2, 3], [1, 2, 3]]
----------
[4, 3, 2, 1, 1, 2, 3, 3, 4, 3]
[[1, 2, 3], [3, 4]]
----------
[1, 4, 3]
[]
----------
[1]
[]
----------
[1, 2, 2]
[[1, 2]]
----------
[2, 1]
[]
----------
[0, 0, 0, 0, 2, 5, 5, 8]
[]
----------
Some explanation. If you "unroll" the code, the logic is more apparant :
from itertools import groupby
def f(i):
return L[i]==L[i-1]+1
def igroups(L):
monotonic_states = [(k,list(g)) for (k, g) in groupby(range(1, len(L)), f)]
increasing_items_indices = [I for (v,I) in monotonic_states if v]
print("\nincreasing_items_indices ->", increasing_items_indices, '\n')
full_increasing_items= [[L[I[0]-1]]+[L[i] for i in I] for I in increasing_items_indices]
return full_increasing_items
L= [2, 8, 4, 5, 6, 7, 8, 5, 9, 10, 11, 12, 25, 26, 27, 42, 41]
print(L)
print(igroups(L))
outputting :
[2, 8, 4, 5, 6, 7, 8, 5, 9, 10, 11, 12, 25, 26, 27, 42, 41]
increasing_items_indices -> [[3, 4, 5, 6], [9, 10, 11], [13, 14]]
[[4, 5, 6, 7, 8], [9, 10, 11, 12], [25, 26, 27]]
We need a key function f that compares an item with the preceding one in the given list. Now, the important point is that the groupby function with the key function f provides a tuple (k, S) where S represents adjacent indices from the initial list and where the state of f is constant, the state being given by the value of k: if k is True, then S represents increasing (by 1) items indices else non-increasing items indices. (in fact, as the example above shows, the list S is incomplete and lacks the first item).
I also made some random tests with one million items lists : igroups function returns always the correct response but is 4 times slower than a naive implementation! Simpler is easier and faster ;)
Thanks alvas for your question, it gives me a lot of fun!
A (really) simple implementation:
x = [7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
result = []
current = x[0]
temp = []
for i in xrange(1, len(x)):
if (x[i] - current == 1):
temp.append( x[i] )
else:
if (len(temp) > 1):
result.append(temp)
temp = [ x[i] ]
current = x[i]
result.append(temp)
And you will get [ [7, 8, 9, 10], [0, 1, 2, 3, 4, 5] ]. From there, you can get the number of increasing numbers by [ len(x) for x in result ] and the total number of numbers sum( len(x) for x in result).
I think this works. It's not fancy but it's simple. It constructs a start list sl and an end list el, which should always be the same length, then uses them to index into x:
def igroups(x):
sl = [i for i in range(len(x)-1)
if (x == 0 or x[i] != x[i-1]+1) and x[i+1] == x[i]+1]
el = [i for i in range(1, len(x))
if x[i] == x[i-1]+1 and (i == len(x)-1 or x[i+1] != x[i]+1)]
return [x[sl[i]:el[i]+1] for i in range(len(sl))]
Late answer, but a simple implementation, generalized to take a predicate so that it need not necessarily be increasing numbers, but could readily be any relationship between the two numbers.
def group_by(lst, predicate):
result = [[lst[0]]]
for i, x in enumerate(lst[1:], start=1):
if not predicate(lst[i - 1], x):
result.append([x])
else:
result[-1].append(x)
return list(filter(lambda lst: len(lst) > 1, result))
Testing this:
>>> group_by([1,2,3,4, 7, 1, 0, 2], lambda x, y: x < y)
[[1, 2, 3, 4, 7], [0, 2]]
>>> group_by([1,2,3,4, 7, 1, 0, 2], lambda x, y: x > y)
[[7, 1, 0]]
>>> group_by([1,2,3,4, 7, 1, 0, 0, 2], lambda x, y: x < y)
[[1, 2, 3, 4, 7], [0, 2]]
>>> group_by([1,2,3,4, 7, 1, 0, 0, 2], lambda x, y: x > y)
[[7, 1, 0]]
>>> group_by([1,2,3,4, 7, 1, 0, 0, 2], lambda x, y: x >= y)
[[7, 1, 0, 0]]
>>>
And now we can easily specialize this:
>>> def ascending_groups(lst):
... return group_by(lst, lambda x, y: x < y)
...
>>> ascending_groups([1,2,3,4, 7, 1, 0, 0, 2])
[[1, 2, 3, 4, 7], [0, 2]]
>>>

Fixed length sequence repeats with increments in python

Looking for a pythonic way to repeat sequences of fixed length while incrementing the sequence digits until max length is reached.
As of now, the code uses while loop and four variables one being the list itself to complete the logic as below,
l = []
i, repeat, max_len = 0, 3, 20
while True:
if len(l) + repeat <= max_len:
l.extend([i] * repeat)
else:
repeat = max_len - len(l)
l.extend([i] * repeat)
break
i += 1
The above code produces
l = [0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6]
Thus, repeating a fixed length sequence of 3 units until the maximum length of 20 is reached (omit any last digits of sequence outside max_len permitted)
Is there a pythonic way of doing the same?
How about this one:
[int(i/repeat) for i in range(max_len)]
Well this will make exactly your list using list comprehension.
l = [i//(repeat) for i in range(max_len)]
# [0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6]
But I think there is a bug in the original function because there should be three 5s at the end.
l = [i//(repeat) for i in range(max_len//repeat * repeat)]
# [0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 5]
With list comprehensions
>>> i, repeat, max_len = 0, 3, 20
>>> q = max_len//repeat
>>> [x for x in range(i, i+q) for y in range(repeat)]+[i+q for y in range(max_len%repeat)]
[0, 0, 0, 1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6]
With a different starting value:
>>> i = 5
>>> [x for x in range(i, i+q) for y in range(repeat)]+[i+q for y in range(max_len%repeat)]
[5, 5, 5, 6, 6, 6, 7, 7, 7, 8, 8, 8, 9, 9, 9, 10, 10, 10, 11, 11]

Finding groups of increasing numbers in a list

The aim is to find groups of increasing/monotonic numbers given a list of integers. Each item in the resulting group must be of a +1 increment from the previous item
Given an input:
x = [7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
I need to find groups of increasing numbers and achieve:
increasing_numbers = [(7,8,9,10), (0,1,2,3,4,5)]
And eventually also the number of increasing numbers:
len(list(chain(*increasing_numbers)))
And also the len of the groups:
increasing_num_groups_length = [len(i) for i in increasing_numbers]
I have tried the following to get the number of increasing numbers:
>>> from itertools import tee, chain
>>> def pairwise(iterable):
... a, b = tee(iterable)
... next(b, None)
... return zip(a, b)
...
>>> x = [8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6]
>>> set(list(chain(*[(i,j) for i,j in pairwise(x) if j-1==i])))
set([1, 2, 3, 4, 5, 6, 8, 9, 10, 11])
>>> len(set(list(chain(*[(i,j) for i,j in pairwise(x) if j-1==i]))))
10
But I'm unable to keep the order and the groups of increasing numbers.
How can I achieve the increasing_numbers groups of integer tuples and also the increasing_num_groups_length?
Also, is there a name for such/similar problem?
EDITED
I've came up with this solution but it's super verbose and I'm sure there's an easier way to achieve the increasing_numbers output:
>>> from itertools import tee, chain
>>> def pairwise(iterable):
... a, b = tee(iterable)
... next(b, None)
... return zip(a, b)
...
>>> x = [8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6]
>>> boundary = iter([0] + [i+1 for i, (j,k) in enumerate(pairwise(x)) if j+1!=k] + [len(x)])
>>> [tuple(x[i:next(boundary)]) for i in boundary]
[(8, 9, 10, 11), (1, 2, 3, 4, 5, 6)]
Is there a more pythonic / less verbose way to do this?
Another input/output example:
[in]:
[17, 17, 19, 20, 21, 22, 0, 1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13,
14, 14, 14, 28, 29, 30, 31, 32, 33, 34, 35, 36, 40]
[out]:
[(19, 20, 21, 22), (0, 1, 2), (4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14),
(28, 29, 30, 31, 32, 33, 34, 35, 36)]
EDIT:
Here's a code-golf solution (142 characters):
def f(x):s=[0]+[i for i in range(1,len(x)) if x[i]!=x[i-1]+1]+[len(x)];return [x[j:k] for j,k in [s[i:i+2] for i in range(len(s)-1)] if k-j>1]
Expanded version:
def igroups(x):
s = [0] + [i for i in range(1, len(x)) if x[i] != x[i-1] + 1] + [len(x)]
return [x[j:k] for j, k in [s[i:i+2] for i in range(len(s)-1)] if k - j > 1]
Commented version:
def igroups(x):
# find the boundaries where numbers are not consecutive
boundaries = [i for i in range(1, len(x)) if x[i] != x[i-1] + 1]
# add the start and end boundaries
boundaries = [0] + boundaries + [len(x)]
# take the boundaries as pairwise slices
slices = [boundaries[i:i + 2] for i in range(len(boundaries) - 1)]
# extract all sequences with length greater than one
return [x[start:end] for start, end in slices if end - start > 1]
Original solution:
Not sure whether this counts as "pythonic" or "not too verbose":
def igroups(iterable):
items = iter(iterable)
a, b = None, next(items, None)
result = [b]
while b is not None:
a, b = b, next(items, None)
if b is not None and a + 1 == b:
result.append(b)
else:
if len(result) > 1:
yield tuple(result)
result = [b]
print(list(igroups([])))
print(list(igroups([0, 0, 0])))
print(list(igroups([7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5])))
print(list(igroups([8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6])))
print(list(igroups([9, 1, 2, 3, 1, 1, 2, 3, 5])))
Output:
[]
[]
[(7, 8, 9, 10), (0, 1, 2, 3, 4, 5)]
[(8, 9, 10, 11), (1, 2, 3, 4, 5, 6)]
[(1, 2, 3), (1, 2, 3)]
A couple of different ways using itertools and numpy:
from itertools import groupby, tee, cycle
x = [17, 17, 19, 20, 21, 22, 0, 1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 14, 14, 28, 29, 30, 31, 32, 33, 34, 35,
36, 1, 2, 3, 4,34,54]
def sequences(l):
x2 = cycle(l)
next(x2)
grps = groupby(l, key=lambda j: j + 1 == next(x2))
for k, v in grps:
if k:
yield tuple(v) + (next((next(grps)[1])),)
print(list(sequences(x)))
[(19, 20, 21, 22), (0, 1, 2), (4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14), (28, 29, 30, 31, 32, 33, 34, 35, 36), (1, 2, 3, 4)]
Or using python3 and yield from:
def sequences(l):
x2 = cycle(l)
next(x2)
grps = groupby(l, key=lambda j: j + 1 == next(x2))
yield from (tuple(v) + (next((next(grps)[1])),) for k,v in grps if k)
print(list(sequences(x)))
Using a variation of my answer here with numpy.split :
out = [tuple(arr) for arr in np.split(x, np.where(np.diff(x) != 1)[0] + 1) if arr.size > 1]
print(out)
[(19, 20, 21, 22), (0, 1, 2), (4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14), (28, 29, 30, 31, 32, 33, 34, 35, 36), (1, 2, 3, 4)]
And similar to ekhumoro's answer:
def sequences(x):
it = iter(x)
prev, temp = next(it), []
while prev is not None:
start = next(it, None)
if prev + 1 == start:
temp.append(prev)
elif temp:
yield tuple(temp + [prev])
temp = []
prev = start
To get the length and the tuple:
def sequences(l):
x2 = cycle(l)
next(x2)
grps = groupby(l, key=lambda j: j + 1 == next(x2))
for k, v in grps:
if k:
t = tuple(v) + (next(next(grps)[1]),)
yield t, len(t)
def sequences(l):
x2 = cycle(l)
next(x2)
grps = groupby(l, lambda j: j + 1 == next(x2))
yield from ((t, len(t)) for t in (tuple(v) + (next(next(grps)[1]),)
for k, v in grps if k))
def sequences(x):
it = iter(x)
prev, temp = next(it), []
while prev is not None:
start = next(it, None)
if prev + 1 == start:
temp.append(prev)
elif temp:
yield tuple(temp + [prev]), len(temp) + 1
temp = []
prev = start
Output will be the same for all three:
[((19, 20, 21, 22), 4), ((0, 1, 2), 3), ((4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14), 11)
, ((28, 29, 30, 31, 32, 33, 34, 35, 36), 9), ((1, 2, 3, 4), 4)]
I think the most maintainable solution would be to make it simple:
def group_by(l):
res = [[l[0]]]
for i in range(1, len(l)):
if l[i-1] < l[i]:
res[-1].append(l[i])
else:
res.append([l[i]])
return res
This solution does not filter out single element sequences, but it can be easily implemented. Additionally, this has O(n) complexity. And you can make it an generator as well if you want.
By maintainable I mean code that is not an one-liner of 300 characters, with some convoluted expressions. Then maybe you would want to use Perl :). At least you will how the function behaves one year later.
>>> x = [7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
>>> print(group_by(x))
[[7, 8, 9, 10], [6], [0, 1, 2, 3, 4, 5]]
If two consecutive numbers are increasing by one I form a list (group) of tuples of those numbers.
When non-increasing and if the list (group) is non-empty, I unpack it and zip again to rebuild the pair of sequence which were broken by the zip. I use set comprehension to eliminate duplicate numbers.
def extract_increasing_groups(seq):
seq = tuple(seq)
def is_increasing(a,b):
return a + 1 == b
def unzip(seq):
return tuple(sorted({ y for x in zip(*seq) for y in x}))
group = []
for a,b in zip(seq[:-1],seq[1:]):
if is_increasing(a,b):
group.append((a,b))
elif group:
yield unzip(group)
group = []
if group:
yield unzip(group)
if __name__ == '__main__':
x = [17, 17, 19, 20, 21, 22, 0, 1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 14, 14, 28, 29, 30, 31, 32, 33, 34, 35, 36, 40]
for group in extract_increasing_groups(x):
print(group)
Simpler one using set;
from collections import namedtuple
from itertools import islice, tee
def extract_increasing_groups(iterable):
iter1, iter2 = tee(iterable)
iter2 = islice(iter2,1,None)
is_increasing = lambda a,b: a + 1 == b
Igroup = namedtuple('Igroup','group, len')
group = set()
for pair in zip(iter1, iter2):
if is_increasing(*pair):
group.update(pair)
elif group:
yield Igroup(tuple(sorted(group)),len(group))
group = set()
if group:
yield Igroup(tuple(sorted(group)), len(group))
if __name__ == '__main__':
x = [17, 17, 19, 20, 21, 22, 0, 1, 2, 2, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 14, 14, 28, 29, 30, 31, 32, 33, 34, 35, 36, 40]
total = 0
for group in extract_increasing_groups(x):
total += group.len
print('Group: {}\nLength: {}'.format(group.group, group.len))
print('Total: {}'.format(total))
def igroups(L):
R=[[]]
[R[-1].append(L[i]) for i in range(len(L)) if (L[i-1]+1==L[i] if L[i-1]+1==L[i] else R.append([L[i]]))]
return [P for P in R if len(P)>1]
tests=[[],
[0, 0, 0],
[7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5],
[8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6],
[9, 1, 2, 3, 1, 1, 2, 3, 5],
[4,3,2,1,1,2,3,3,4,3],
[1, 4, 3],
[1],
[1,2],
[2,1]
]
for L in tests:
print(L)
print(igroups(L))
print("-"*10)
outputting the following:
[]
[]
----------
[0, 0, 0]
[]
----------
[7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
[[7, 8, 9, 10], [0, 1, 2, 3, 4, 5]]
----------
[8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6]
[[8, 9, 10, 11], [1, 2, 3, 4, 5, 6]]
----------
[9, 1, 2, 3, 1, 1, 2, 3, 5]
[[1, 2, 3], [1, 2, 3]]
----------
[4, 3, 2, 1, 1, 2, 3, 3, 4, 3]
[[1, 2, 3], [3, 4]]
----------
[1, 4, 3]
[]
----------
[1]
[]
----------
[1, 2]
[[1, 2]]
----------
[2, 1]
[]
----------
EDIT
My first attemp using itertools.groupby was a fail, sorry for that.
With itertools.groupby, the problem of partionning a list of integers L in sublists of adjacent and increasing consecutive items from L can be done with a one-liner. Nevertheless I don't know how pythonic it can be considered ;)
Here is the code with some simple tests:
[EDIT : now subsequences are increasing by 1, I missed this point the first time.]
from itertools import groupby
def f(i):
return L[i-1]+1==L[i]
def igroups(L):
return [[L[I[0]-1]]+[L[i] for i in I] for I in [I for (v,I) in [(k,[i for i in list(g)]) for (k, g) in groupby(range(1, len(L)), f)] if v]]
outputting:
tests=[
[0, 0, 0, 0],
[7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5],
[8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6],
[9, 1, 2, 3, 1, 1, 2, 3, 5],
[4,3,2,1,1,2,3,3,4,3],
[1, 4, 3],
[1],
[1,2, 2],
[2,1],
[0, 0, 0, 0, 2, 5, 5, 8],
]
for L in tests:
print(L)
print(igroups(L))
print('-'*10)
[0, 0, 0, 0]
[]
----------
[7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
[[7, 8, 9, 10], [0, 1, 2, 3, 4, 5]]
----------
[8, 9, 10, 11, 7, 1, 2, 3, 4, 5, 6]
[[8, 9, 10, 11], [1, 2, 3, 4, 5, 6]]
----------
[9, 1, 2, 3, 1, 1, 2, 3, 5]
[[1, 2, 3], [1, 2, 3]]
----------
[4, 3, 2, 1, 1, 2, 3, 3, 4, 3]
[[1, 2, 3], [3, 4]]
----------
[1, 4, 3]
[]
----------
[1]
[]
----------
[1, 2, 2]
[[1, 2]]
----------
[2, 1]
[]
----------
[0, 0, 0, 0, 2, 5, 5, 8]
[]
----------
Some explanation. If you "unroll" the code, the logic is more apparant :
from itertools import groupby
def f(i):
return L[i]==L[i-1]+1
def igroups(L):
monotonic_states = [(k,list(g)) for (k, g) in groupby(range(1, len(L)), f)]
increasing_items_indices = [I for (v,I) in monotonic_states if v]
print("\nincreasing_items_indices ->", increasing_items_indices, '\n')
full_increasing_items= [[L[I[0]-1]]+[L[i] for i in I] for I in increasing_items_indices]
return full_increasing_items
L= [2, 8, 4, 5, 6, 7, 8, 5, 9, 10, 11, 12, 25, 26, 27, 42, 41]
print(L)
print(igroups(L))
outputting :
[2, 8, 4, 5, 6, 7, 8, 5, 9, 10, 11, 12, 25, 26, 27, 42, 41]
increasing_items_indices -> [[3, 4, 5, 6], [9, 10, 11], [13, 14]]
[[4, 5, 6, 7, 8], [9, 10, 11, 12], [25, 26, 27]]
We need a key function f that compares an item with the preceding one in the given list. Now, the important point is that the groupby function with the key function f provides a tuple (k, S) where S represents adjacent indices from the initial list and where the state of f is constant, the state being given by the value of k: if k is True, then S represents increasing (by 1) items indices else non-increasing items indices. (in fact, as the example above shows, the list S is incomplete and lacks the first item).
I also made some random tests with one million items lists : igroups function returns always the correct response but is 4 times slower than a naive implementation! Simpler is easier and faster ;)
Thanks alvas for your question, it gives me a lot of fun!
A (really) simple implementation:
x = [7, 8, 9, 10, 6, 0, 1, 2, 3, 4, 5]
result = []
current = x[0]
temp = []
for i in xrange(1, len(x)):
if (x[i] - current == 1):
temp.append( x[i] )
else:
if (len(temp) > 1):
result.append(temp)
temp = [ x[i] ]
current = x[i]
result.append(temp)
And you will get [ [7, 8, 9, 10], [0, 1, 2, 3, 4, 5] ]. From there, you can get the number of increasing numbers by [ len(x) for x in result ] and the total number of numbers sum( len(x) for x in result).
I think this works. It's not fancy but it's simple. It constructs a start list sl and an end list el, which should always be the same length, then uses them to index into x:
def igroups(x):
sl = [i for i in range(len(x)-1)
if (x == 0 or x[i] != x[i-1]+1) and x[i+1] == x[i]+1]
el = [i for i in range(1, len(x))
if x[i] == x[i-1]+1 and (i == len(x)-1 or x[i+1] != x[i]+1)]
return [x[sl[i]:el[i]+1] for i in range(len(sl))]
Late answer, but a simple implementation, generalized to take a predicate so that it need not necessarily be increasing numbers, but could readily be any relationship between the two numbers.
def group_by(lst, predicate):
result = [[lst[0]]]
for i, x in enumerate(lst[1:], start=1):
if not predicate(lst[i - 1], x):
result.append([x])
else:
result[-1].append(x)
return list(filter(lambda lst: len(lst) > 1, result))
Testing this:
>>> group_by([1,2,3,4, 7, 1, 0, 2], lambda x, y: x < y)
[[1, 2, 3, 4, 7], [0, 2]]
>>> group_by([1,2,3,4, 7, 1, 0, 2], lambda x, y: x > y)
[[7, 1, 0]]
>>> group_by([1,2,3,4, 7, 1, 0, 0, 2], lambda x, y: x < y)
[[1, 2, 3, 4, 7], [0, 2]]
>>> group_by([1,2,3,4, 7, 1, 0, 0, 2], lambda x, y: x > y)
[[7, 1, 0]]
>>> group_by([1,2,3,4, 7, 1, 0, 0, 2], lambda x, y: x >= y)
[[7, 1, 0, 0]]
>>>
And now we can easily specialize this:
>>> def ascending_groups(lst):
... return group_by(lst, lambda x, y: x < y)
...
>>> ascending_groups([1,2,3,4, 7, 1, 0, 0, 2])
[[1, 2, 3, 4, 7], [0, 2]]
>>>

Extract a value out of n

How do you extract a value out of n of a list in python ?
For example :
n = 3
l = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
And I would like to get
[0, 3, 6, 9]
I know I can do this with a for, but is there any more pythonic and short way ?
You can do a simple list comprehension
>>> n = 3
>>> l = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
>>> [i for i in l if i%n==0]
[0, 3, 6, 9]
If your list is always like that, then you can use strides
>>> l[::3]
[0, 3, 6, 9]
Tip
Use range to generate lists like that
>>> range(10)
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
Using slicing would be the most pythonic way:
In [1]: n = 3
In [2]: l = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
In [3]: l[::n]
Out[3]: [0, 3, 6, 9]
Use a slice with stride:
l[::n]
Demo:
>>> n = 3
>>> l = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
>>> l[::n]
[0, 3, 6, 9]
print l[::3] # slice with step n = 3

convert an int to list of individual digitals more faster?

All,
I want define an int(987654321) <=> [9, 8, 7, 6, 5, 4, 3, 2, 1] convertor, if the length of int number < 9, for example 10 the list will be [0,0,0,0,0,0,0,1,0]
, and if the length > 9, for example 9987654321 , the list will be [9, 9, 8, 7, 6, 5, 4, 3, 2, 1]
>>> i
987654321
>>> l
[9, 8, 7, 6, 5, 4, 3, 2, 1]
>>> z = [0]*(len(unit) - len(str(l)))
>>> z.extend(l)
>>> l = z
>>> unit
[100000000, 10000000, 1000000, 100000, 10000, 1000, 100, 10, 1]
>>> sum([x*y for x,y in zip(l, unit)])
987654321
>>> int("".join([str(x) for x in l]))
987654321
>>> l1 = [int(x) for x in str(i)]
>>> z = [0]*(len(unit) - len(str(l1)))
>>> z.extend(l1)
>>> l1 = z
>>> l1
[9, 8, 7, 6, 5, 4, 3, 2, 1]
>>> a = [i//x for x in unit]
>>> b = [a[x] - a[x-1]*10 for x in range(9)]
>>> if len(b) = len(a): b[0] = a[0] # fix the a[-1] issue
>>> b
[9, 8, 7, 6, 5, 4, 3, 2, 1]
I tested above solutions but found those may not faster/simple enough than I want and may have a length related bug inside, anyone may share me a better solution for this kinds convertion?
Thanks!
Maybe I am missing something, but shouldn't this be enough (without value checking)?
def int_to_list(i):
return [int(x) for x in str(i).zfill(9)]
def list_to_int(l):
return int("".join(str(x) for x in l))
Reference: str.zfill
And what about :
def int_to_list(num)
return list ("%010d" % num)
def convert(number):
stringified_number = '%s' % number
if len(stringified_number) < 9:
stringified_number = stringified_number.zfill(9)
return [int(c) for c in stringified_number]
>>> convert(10)
[0, 0, 0, 0, 0, 0, 0, 1, 0]
>>> convert(987654321)
[9, 8, 7, 6, 5, 4, 3, 2, 1]
To place an integer of any length into a list in sequence by integer digit -
a = 123456789123456789123456789123456789123456789123456789
j = len('{}'.format(a))
b = [0 for i in range(j)]
c = 0
while j > 0:
b [c] = a % 10**j // 10**(j-1)
j = j-1
c = c + 1
print(b)
output -
[1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1, 2, 3, 4, 5, 6, 7, 8, 9]
you can put the condition on j for the alternate assignment to b.

Categories