Str.Contains show only True values

Question

I am using str.contains on a large dataframe and I need a way such that str.contains returns the records where my str.contains function is True. (the dataframe is several thousand lines long and I am looking for 8 true responses).

Thanks!

aa = filtered_to_df.body.str.contains('AA')

aa.head(10)  
Out[312]:
15864    False  
18040    False  
22576    False  
28092    False  
32800    False  
33236    False   
38027    False  
41222    False  
46647    False  
87645    False  
Name: body, dtype: bool

added - I was thinking is there a way to use loc on the series? — Tony, Dec 04 '17 at 03:54
I just added it: aa = filtered_to_df.body.str.contains('AA') — Tony, Dec 04 '17 at 03:55
If you want negation use tilde `~` like `~aa` or `aa = ~filtered.....` — Bharath M Shetty, Dec 04 '17 at 03:56
This doesn't appear to be working it still shows all False responses: filtered_to_df.body.astype(str).str.contains('AA') 15864 False 18040 False 22576 False 28092 False 32800 False 33236 False 38027 False — Tony, Dec 04 '17 at 03:57
Unfortunately that appears to be inverting the values? False now = True, True now = False? the values I am looking for are still buried in the data abbreviation — Tony, Dec 04 '17 at 04:00
But you are supposed to add sample of your actual data we don't want the result you are getting. There might be 100 reasons why you are getting false. So add the sample of actual data — Bharath M Shetty, Dec 04 '17 at 04:10

score 2 · Answer 1 · answered Dec 04 '17 at 05:50

important distinction: str.contains does not actually filter your dataframe or series, it just returns a boolean vector of the same dimension as the series you applied it on.

e.g: if you have a series like this:

my_series = pd.Series(['hello world', 'hello', 'world'])

print(my_series)

0    hello world
1          hello
2          world
dtype: object

using str.contains("hello") on this will return a series of size 3 since it will give you True / False for every cell in the series -- does that cell contain the word "hello"?

my_series = pd.Series(['hello world', 'hello', 'world'])

print(my_series.str.contains("hello"))

0     True
1     True
2    False
dtype: bool

to actually filter the dataframe or series, you need to wrap around it with a slicing operation.

my_series[my_series.str.contains("hello")]

0    hello world
1          hello
dtype: object

Str.Contains show only True values

1 Answers1

Related