I have this string
"id=IAfpK, val=58, id=WNVdi, val=64, id=jp9zt, val=47"
. I want to create a pandas data frame out of it. Any ideas how to do so?
Asked
Active
Viewed 112 times
-1
-
`import pandas as pd` `string = "id=IAfpK, val=58, id=WNVdi, val=64, id=jp9zt, val=47"` `df = pd.DataFrame({'String' : [string]})` `df` – David Erickson Aug 21 '20 at 23:32
-
@DavidErickson i want a dataframe with two columns and of length 3 as there are 3 key, val pairs – jasraj bedi Aug 21 '20 at 23:35
-
makes sense. These details should be in your question if you can modify it and make it clear exactly what you want. – David Erickson Aug 21 '20 at 23:37
-
@DavidErickson If there is no duplicate why to vote to close it ! – jasraj bedi Aug 21 '20 at 23:38
-
I retracted my vote, since it is not as simple as the code I copy and pasted above after you have explained it. – David Erickson Aug 21 '20 at 23:39
1 Answers
1
import pandas as pd
data = "id=IAfpK, val=58, id=WNVdi, val=64, id=jp9zt, val=47"
data = [*map(lambda x: x.split('='), data.split(', '))]
ids = [*filter(lambda x: x[0] == 'id', data)]
vals = [*filter(lambda x: x[0] == 'val', data)]
df = pd.DataFrame.from_dict(
data={
'id': [*map(lambda x: x[1], ids)],
'val': [*map(lambda x: x[1], vals)]
}
)
Dimitri Sifoua
- 490
- 5
- 17
-
Useful only for a very short data and is just unnecessarily complex. Not everything has to be a `map()`, `lambda` or `filter()`. Use it where necessary/useful and always use `timeit` when using too many of them. In my experience it's harder to read and worse code overall (but has its uses). Instead: ```>>> def better(data): ... from collections import defaultdict as ddict ... out = ddict(list) ... data = [it.split("=") for it in [item.strip() for item in data.split(",")]] ... for key, val in data: ... out[key].append(val) ... return out``` – Peter Badida Aug 22 '20 at 00:06
-
Sorry, the comments can't format it properly. Just add a newline on each `...`. – Peter Badida Aug 22 '20 at 00:08