Python Forum

Hi,

I am working with a pandas dataframe which has a date column, called Occurrence Year. I am trying to only get the year for that column (drop the month and day) so that I can pull the data that is for the year 2015.

This is what the original dataframe looks like...


And this is what I tried:

Which gives me NaN values in my column.

So I tried using the .apply code:

NYCrime['Year'] = NYCrime['Occurrence Year'].apply(lambda x: x[:-4])

which then gave me the error:

Error:
TypeError                                 Traceback (most recent call last)
<ipython-input-47-ef12f8aeb1c4> in <module>
      2 #NYCrime['Occurrence Year'].astype(str)
      3 #NYCrime['Occurrence Year'].dtype
----> 4 NYCrime['Year'] = NYCrime['Occurrence Year'].apply(lambda x: x[:-4])
      5 
      6 #IGNORE TCrimedf['Neighbourhood'] = TCrimedf['Neighbourhood'].str.rstrip('(0123456789)')

~/conda/envs/python/lib/python3.6/site-packages/pandas/core/series.py in apply(self, func, convert_dtype, args, **kwds)
   3846             else:
   3847                 values = self.astype(object).values
-> 3848                 mapped = lib.map_infer(values, f, convert=convert_dtype)
   3849 
   3850         if len(mapped) and isinstance(mapped[0], Series):

pandas/_libs/lib.pyx in pandas._libs.lib.map_infer()

<ipython-input-47-ef12f8aeb1c4> in <lambda>(x)
      2 #NYCrime['Occurrence Year'].astype(str)
      3 #NYCrime['Occurrence Year'].dtype
----> 4 NYCrime['Year'] = NYCrime['Occurrence Year'].apply(lambda x: x[:-4])
      5 
      6 #IGNORE TCrimedf['Neighbourhood'] = TCrimedf['Neighbourhood'].str.rstrip('(0123456789)')

TypeError: 'float' object is not subscriptable

So I tried to use the .astype to change the values to string...

NYCrime['Occurrence Year'].astype(str)

and run the .apply code again but still got the same error.

I checked the data type of the columns (

NYCrime['Occurrence Year'].dtype

)
and got dtype('O')

Can someone help me, please? I'm at a loss at how to figure this out

split returns a tuple. When you did split date using '/' it returned ('12', '31', '2015'). You looked at the wrong index.

Your lambda didn't work because the date is not a string. The error message says it is a float.

I am pretty sure pandas has built in functions for extracting the information you want. Read about pandas date/time functions.

Convert so Occurrence Year is a pandas date-time object,then can extract dates for 2015.
Example.

import pandas as pd
from io import StringIO

data = StringIO('''\
occurrence_year,values
10/31/2014,7.0
10/31/2015,2.0
12/31/2015,3.0
10/31/2016,9.0''')

df = pd.read_csv(data, sep=',')
print(df)

Output:
  occurrence_year  values
0      10/31/2014     7.0
1      10/31/2015     2.0
2      12/31/2015     3.0
3      10/31/2016     9.0

# Check types
df.dtypes

occurrence_year     object
values             float64
dtype: object

# Fix date type
df['occurrence_year'] = pd.to_datetime(df['occurrence_year'])
df.dtypes

occurrence_year    datetime64[ns]
values                    float64
dtype: object

# Extract dates for 2015
df[df['occurrence_year'].dt.year == 2015]


occurrence_year	values
1	2015-10-31	2.0
2	2015-12-31	3.0