Filtering a list in python using a function of other values in the list.

Question

counts = misc. array

    for i in counts:
        if i < "(sum of the two integers next to i)/4"
            return -1

What I am trying to do is set all values in the list which are less than half the average of their two neighbours to -1. i.e. for the list [10., 20., 10., 25., 30.] if i=3 then I want to see if 10 is less than (20+25)/4 and since that is 45/4 > 10 I want to then set 10 equal to -1. I will then filter for all values in the list >= 0.

This is further complicated by the fact that you should start with i+1 because this process doesn't work with the first value and then you should end with -1 because this process doesn't work with the last value. Any help would be greatly appreciated.

Community · Accepted Answer · 2017-05-23 12:21:06Z

1

You could borrow the sliding window generator from this answer, and put it to use like so:

out = [y if y >= (x+z)/4. else -1 for x,y,z in window(l, 3)]

where l is your list.

This chops off the first and the last element. It would be easy to do something different to them (e.g. set to -1) if you so wish.

edited May 23, 2017 at 12:21

CommunityBot

11 silver badge

answered Mar 16, 2013 at 19:05

NPE

503k114 gold badges970 silver badges1k bronze badges

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

1 Comment

catalesia Over a year ago

how does that help user(new_user_number) who just about asked his first question? Not that my answer is the greatest...

ConfusedProgrammer · Accepted Answer · 2013-03-16 19:05:46Z

0

Does this do what you want (assuming values are in list data)?

filtered = [data[i] for i in range(1,len(data)-1) if data[i] >= (data[i+1]+data[i-1])/4]

answered Mar 16, 2013 at 19:05

ConfusedProgrammer

6161 gold badge5 silver badges14 bronze badges

Comments

uselpa · Accepted Answer · 2013-03-16 19:41:55Z

0

I have a solution that doesn't look too elegant, but I believe is correct and understandable. The filtering is included, so it will also work if -1 is an allowed value.

I use 3 variables (a,b and i), and 3 flag variables (fa, fb and fi) which show if the value needs to be included in the result set or not.

def test(lst):
    a, fa, b, fb, res  = None, False, None, False, []
    for i in lst:
        if a is not None:
            fi = b < (a + i)/4
        else:
            fi = True
        if fa:
            res.append(a)
        a, fa = b, fb
        b, fb = i, fi
    if fa:
        res.append(a)
    res.append(b)
    return res

print(test([10., 20., 10., 25., 30.]))
=> [10.0, 20.0, 25.0, 30.0]

answered Mar 16, 2013 at 19:41

uselpa

19k2 gold badges36 silver badges53 bronze badges

Comments

catalesia · Accepted Answer · 2013-03-16 23:05:32Z

Like that:

result = [(x+1 if counts[x] > counts[x-1]/counts[x+1]/4 else -1) for x in range(len(counts[1:-1]))]

- x is a idex in counts --> for x in range(len(counts[1:-1])
- x+1 because we know counts[0] and count[-1] are out. Hence to achieve result one must adjust it.
- if count[x] > counts[x-1]/counts[x+1]/4 else -1 test, returns the result if result is True, otherwise -1
note the > 0, it's opposite to what you want
- for x in range(len(counts[1:-1]))
because we know first and last value is of no matter
- counts[1:-1] as above, always better to use a copy of a list, thread-safe

This should work, will give you a list of valid (!=-1) and invalid (== -1) results per index in count,

question is, what can you do further with the result?

you can test, e.g.: (mind the adjusted x in result)

 ['N/A'] + [(x if x!= -1 else False) for x in result] + ['N/A']

which gives you a mirror image of counts with either False or True result

Collectives™ on Stack Overflow

Filtering a list in python using a function of other values in the list.

4 Answers 4

1 Comment

Comments

Comments

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

1 Comment

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related