Given a data frame like below
A B C-1 D BTP Type C1 Type C2
0 0 1 0 0 0 NaN [Type B]
1 0 2 1 1 14 [Type B] [Type B]
2 0 3 2 2 28 [Type A] [Type B]
3 0 4 3 3 42 [Type A, Type B] [Type A, Type B]
4 0 5 4 4 56 [Type A] [Type A, Type B]
want to fetch rows with value Type A
for column Type C1
and 42
for column BTP
which should return row index 3.
Tried the following, but gives an error KeyError: False
df.loc[(df['BTP'] == 42) & ('Type A' in df['Type C1'])]
What I'm ultimately trying to do is to fetch row that will match the above condition (which would be a single row) and extract the values for columns B
and C-1
as a dict like {'B_val': 4, 'C_val': 3}
Use, Series.str.join
to join the lists in column Type C1
, then we could be able to use Series.str.contains
on this column to check whether the given string ie Type A
is present in the series or not, finally we can filter the rows of dataframe using the boolean mask
:
mask = df['BTP'].eq(42) & df['Type C1'].str.join('-').str.contains(r'\bType A\b')
df = df[mask]
Result:
# print(df)
A B C-1 D BTP Type C1 Type C2
3 0 4 3 3 42 [Type A, Type B] [Type A, Type B]
You can use
>>> type_a = df['Type C1'].apply(pd.Series).eq('Type A').any(1)
>>> df[df['BTP'].eq(42) & type_a]
A B C-1 D BTP Type C1 Type C2
3 0 4 3 3 42 [Type A, Type B] [Type A, Type B]
I solved this with a custom function to return a list of True/False values for each row, based on whether the list under consideration contains 'Type A' or not.
# Check if elem is present in column 'col'
def has_elem(col, elem):
result = []
for c in col:
if elem in c:
result.append(True)
else:
result.append(False)
return result
# Filter
df.loc[(df['BTP'] == 42) & has_elem(df['Type_C1'], 'Type A'), :]
The reason your code doesn't work is because the 2nd filter clause 'Type A' in df['Type_C1']
looks for membership of the string 'Type A'
in the Series object df['Type_C1']
, and consequently returns False
. Instead, you need to return a sequence of True/False values, 1 for each row in your dataframe.
The technical post webpages of this site follow the CC BY-SA 4.0 protocol. If you need to reprint, please indicate the site URL or the original address.Any question please contact:yoyou2525@163.com.