问题
I have a list of tuples:
[(0.0, 287999.70000000007),
(1.0, 161123.23000000001),
(2.0, 93724.140000000014),
(3.0, 60347.309999999983),
(4.0, 55687.239999999998),
(5.0, 29501.349999999999),
(6.0, 14993.920000000002),
(7.0, 14941.970000000001),
(8.0, 13066.229999999998),
(9.0, 10101.040000000001),
(10.0, 4151.6900000000005),
(11.0, 2998.8899999999999),
(12.0, 1548.9300000000001),
(15.0, 1595.54),
(16.0, 1435.98),
(17.0, 1383.01)]
As can be seen, there are missing indexes (13 and 14). I want to fill the missing indexes with zeros:
[(0.0, 287999.70000000007),
(1.0, 161123.23000000001),
(2.0, 93724.140000000014),
(3.0, 60347.309999999983),
(4.0, 55687.239999999998),
(5.0, 29501.349999999999),
(6.0, 14993.920000000002),
(7.0, 14941.970000000001),
(8.0, 13066.229999999998),
(9.0, 10101.040000000001),
(10.0, 4151.6900000000005),
(11.0, 2998.8899999999999),
(12.0, 1548.9300000000001),
(13.0, 0),
(14.0, 0),
(15.0, 1595.54),
(16.0, 1435.98),
(17.0, 1383.01)]
I did something ugly with for loop (I didn't add it cause I don't think it will contribute to anything...), but I was wondering is there any elegant way to resolve this problem? (maybe 3-4 lines with list comprehension).
回答1:
Just a straight for loop is probably easier than a list comprehension:
data = [(0.0, 287999.70000000007),
(1.0, 161123.23000000001),
(2.0, 93724.140000000014),
(3.0, 60347.309999999983),
(4.0, 55687.239999999998),
(5.0, 29501.349999999999),
(6.0, 14993.920000000002),
(7.0, 14941.970000000001),
(8.0, 13066.229999999998),
(9.0, 10101.040000000001),
(10.0, 4151.6900000000005),
(11.0, 2998.8899999999999),
(12.0, 1548.9300000000001),
(15.0, 1595.54),
(16.0, 1435.98),
(17.0, 1383.01)]
result = []
last = 0.0
for d in data:
while last < d[0]:
result.append((last, 0))
last += 1
result.append(d)
last = d[0]+1
Slightly shorter (and including a list comprehension):
result, last = [], 0.0
for d in data:
result.extend((r,0) for r in range(int(last), int(d[0])))
result.append(d)
last = d[0]+1
回答2:
I've modified your input slightly to use integer values.
Presuming the input is in order. First I work out the highest key in the list. top=in_list[-1][0]
Then convert the input to a dict.
This means I can use get(key[, default]) to return a zero if the key doesn't exist.
Then use a list comprehension with range to go through the possible integers. Needs to be top+1 because range returns the number of elements, and so starting from zero, needs one more.
list=[(0, 287999.70000000007),
(1, 161123.23000000001),
(2, 93724.140000000014),
(3, 60347.309999999983),
(4, 55687.239999999998),
(5, 29501.349999999999),
(6, 14993.920000000002),
(7, 14941.970000000001),
(8, 13066.229999999998),
(9, 1010140000000001),
(10, 4151.6900000000005),
(11, 2998.8899999999999),
(12, 1548.9300000000001),
(15, 1595.54),
(16, 1435.98),
(17, 1383.01)]
top=in_list[-1][0]
in_dict=dict(in_list)
out_list=[ (i,in_dict.get(i,0)) for i in range(top+1)]
print(out_list)
回答3:
You can do it by converting your data to a dictionary, and then retrieving each index with dict.get() so that you can default the value to 0.
Compact version:
def fill_missing_with_zero(collection, fill=0.0):
d = dict(collection)
return [(key, d.get(key, fill)) for key in [float(i) for i in range(int(max(d))+1)]]
For more detail, As follows:
def fill_missing_with_zero(collection, fill=0.0):
d = dict(collection)
highest_index = int(max(d.keys()))
result = []
for i in range(highest_index+1):
key = float(i) # because your "keys" are floats
result.append((key, d.get(key, fill)))
return result
Example:
>>> fill_missing_with_zero(collection))
[(0.0, 287999.70000000007),
(1.0, 161123.23),
(2.0, 93724.14000000001),
(3.0, 60347.30999999998),
(4.0, 55687.24),
(5.0, 29501.35),
(6.0, 14993.920000000002),
(7.0, 14941.970000000001),
(8.0, 13066.229999999998),
(9.0, 10101.04),
(10.0, 4151.6900000000005),
(11.0, 2998.89),
(12.0, 1548.93),
(13.0, 0.0),
(14.0, 0.0),
(15.0, 1595.54),
(16.0, 1435.98),
(17.0, 1383.01)]
来源：https://stackoverflow.com/questions/35840403/python-filling-a-list-of-tuples-with-zeros-in-places-of-missing-indexes