pandas warning with pd.to_datetime - python

Using pandas 0.6.2. I want to change a dataframe to datetime type, here is the dataframe
>>> tt.head()
0 2015-02-01 00:46:28
1 2015-02-01 00:59:56
2 2015-02-01 00:16:27
3 2015-02-01 00:33:45
4 2015-02-01 13:48:29
Name: TS, dtype: object
And I want change each items in tt into datetime type, and get the hour. The code is
for i in tt.index:
tt[i]=pd.to_datetime(tt[i])
and waring is
__main__:2: SettingWithCopyWarning:
A value is trying to be set on a copy of a slice from a DataFrame
See the the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy
Why the warning occurs and how can I deal with it?
If I change one item each time, it works, the code is
>>> tt[1]=pd.to_datetime(tt[1])
>>> tt[1].hour
0

Just do it on the entire Series as to_datetime can operate on array-like args and assign directly to the column:
In [72]:
df['date'] = pd.to_datetime(df['date'])
df.info()
<class 'pandas.core.frame.DataFrame'>
Int64Index: 5 entries, 0 to 4
Data columns (total 1 columns):
date 5 non-null datetime64[ns]
dtypes: datetime64[ns](1)
memory usage: 80.0 bytes
In [73]:
df
Out[73]:
date
index
0 2015-02-01 00:46:28
1 2015-02-01 00:59:56
2 2015-02-01 00:16:27
3 2015-02-01 00:33:45
4 2015-02-01 13:48:29
If you changed your loop to this then it would work:
In [80]:
for i in df.index:
df.loc[i,'date']=pd.to_datetime(df.loc[i, 'date'])
df
Out[80]:
date
index
0 2015-02-01 00:46:28
1 2015-02-01 00:59:56
2 2015-02-01 00:16:27
3 2015-02-01 00:33:45
4 2015-02-01 13:48:29
the code moans because you're operating on potentially a copy of that row on the df and not a view, using the new indexers avoids this ambiguity
EDIT
It looks like you're using an ancient version of pandas, the following should work:
tt[1].apply(lambda x: x.hour)

Related

Extracting YYYY-MM from datetime column

I've a dataframe of this format -
var1 date
A 2017/01/01
A 2017/01/02
...
I want the date to be converted into YYYY-MM format but the df['date'].dtype is object.
How can I remove the day part from date while keeping the data type as datetime?
Expected Output -
A - 2017/01
Thanks
You can't have custom representation for the datetime dtype. But you have the following options:
use strings - you might have any representation (as you wish), but all datetime methods and attributes get lost
use datetime, but set the day part to 1 (as #Kopytok) has already shown.
use period dtype, which still allows you to use some date arithmetic
Demo:
In [207]: df
Out[207]:
var1 date
0 A 2018-12-31
1 A 2017-09-07
2 B 2016-02-29
In [208]: df['new'] = df['date'].dt.to_period('M')
In [209]: df
Out[209]:
var1 date new
0 A 2018-12-31 2018-12
1 A 2017-09-07 2017-09
2 B 2016-02-29 2016-02
In [210]: df.dtypes
Out[210]:
var1 object
date datetime64[ns]
new object
dtype: object
In [211]: df['new'] + 8
Out[211]:
0 2019-08
1 2018-05
2 2016-10
Name: new, dtype: object
It is possible replace every date with the first day of month:
pd.to_datetime(d["date"], format="%Y/%m/%d").apply(lambda x: x.replace(day=1))
Result:
0 2017-01-01
1 2017-01-01

How to find a partial numeric value in column in Pandas?

I have a data frame created with Pandas. It has 3 columns. One of them has the date in the format %Y%m%d%H. I need to find the rows that match a date with the format %Y%m%d.
I tried
df.loc[df["MESS_DATUM"] == 20170807]
which doesn't work. Only when I do
df.loc[df["MESS_DATUM"] == 2017080723]
it works for that single line. But I need the other lines containing the date only (without the hour). I know there is something like .str.cotains(""). Is there something similar for numeric values or a way to use wildcards in the lines above?
We can "integer divide" MESS_DATUM column by 100:
df.loc[df["MESS_DATUM"]//100 == 20170807]
Demo:
In [29]: df
Out[29]:
MESS_DATUM
0 2017080719
1 2017080720
2 2017080721
3 2017080722
4 2017080723
In [30]: df.dtypes
Out[30]:
MESS_DATUM int64
dtype: object
In [31]: df["MESS_DATUM"]//100
Out[31]:
0 20170807
1 20170807
2 20170807
3 20170807
4 20170807
Name: MESS_DATUM, dtype: int64
But I would consider converting it to datetime dtype:
df["MESS_DATUM"] = pd.to_datetime(df["MESS_DATUM"].astype(str), format='%Y%m%d%H')
If df["MESS_DATUM"] is of float dtype, then we can use the following trick:
In [41]: pd.to_datetime(df["MESS_DATUM"].astype(str).str.split('.').str[0],
format='%Y%m%d%H')
Out[41]:
0 2017-08-07 19:00:00
1 2017-08-07 20:00:00
2 2017-08-07 21:00:00
3 2017-08-07 22:00:00
4 2017-08-07 23:00:00
Name: MESS_DATUM, dtype: datetime64[ns]

How can I create a datetime column without 'date' part?

I have a dataframe and there's a column named 'Time' in it like the below(HH:MM:SS:fffff).
>>> df['Time']
0 09:42:29:75284
1 09:42:29:95584
2 09:42:31:15036
3 09:42:35:15138
4 09:42:35:95491
5 09:42:43:55414
6 09:42:45:35866
7 09:42:46:74638
8 09:42:47:35582
9 09:42:47:74774
10 09:42:48:94582
...
Name: Time, Length: 18924, dtype: object
I want to change its type as datetime, in order to make it easier to calculate. Is it possible to change its type, using pandas.to_datetime, as datetime without date?
You can convert it to timedelta64[ns] dtype:
Source DF:
In [164]: df
Out[164]:
Time
0 09:42:29:75284
1 09:42:29:95584
2 09:42:31:15036
3 09:42:35:15138
4 09:42:35:95491
5 09:42:43:55414
6 09:42:45:35866
7 09:42:46:74638
8 09:42:47:35582
9 09:42:47:74774
10 09:42:48:94582
In [165]: df.dtypes
Out[165]:
Time object # <-------- NOTE!
dtype: object
Converted:
In [166]: df.Time = pd.to_timedelta(df.Time.str.replace(r'\:(\d+)$', r'.\1'),
errors='coerce')
In [167]: df
Out[167]:
Time
0 09:42:29.752840
1 09:42:29.955840
2 09:42:31.150360
3 09:42:35.151380
4 09:42:35.954910
5 09:42:43.554140
6 09:42:45.358660
7 09:42:46.746380
8 09:42:47.355820
9 09:42:47.747740
10 09:42:48.945820
In [168]: df.dtypes
Out[168]:
Time timedelta64[ns] # <-------- NOTE!
dtype: object
Please refer python to_datetime documentation.
import pandas as pd
df = pd.DataFrame({'Time': ['09:42:29:75284','09:42:29:95584','09:42:31:15036']})
df
Out[]:
Time
0 09:42:29:75284
1 09:42:29:95584
2 09:42:31:15036
You can convert this into datetime format by specifying format as follows:
pd.to_datetime(df['Time'], format='%H:%M:%S:%f')
Out[]:
0 1900-01-01 09:42:29.752840
1 1900-01-01 09:42:29.955840
2 1900-01-01 09:42:31.150360
Name: Time, dtype: datetime64[ns]
but doing this will also add date 1900-01-01.

is the year-month part of a datetime variable still a time object in Pandas?

consider this
df=pd.DataFrame({'A':['20150202','20150503','20150503'],'B':[3, 3, 1],'C':[1, 3, 1]})
df.A=pd.to_datetime(df.A)
df['month']=df.A.dt.to_period('M')
df
Out[59]:
A B C month
0 2015-02-02 3 1 2015-02
1 2015-05-03 3 3 2015-05
2 2015-05-03 1 1 2015-05
and my month variable is:
df.month
Out[82]:
0 2015-02
1 2015-05
2 2015-05
Name: month, dtype: object
Now if I index my dataset by df.month, it seems that Pandas understands this is a date. In other words, I can draw a plot without having to sort my index first.
But is this actually correct? The dtype object (instead of some datetime format) worries me. Is there a proper date object type for this kind of monthly date?
It is a pandas period object
In [5]: df.month.map(type)
Out[5]:
0 <class 'pandas._period.Period'>
1 <class 'pandas._period.Period'>
2 <class 'pandas._period.Period'>
Name: month, dtype: object

Pandas and csv import into dataframe. How to best to combine date anbd date fields into one

I have a csv file that I am trying to import into pandas.
There are two columns of intrest. date and hour and are the first two cols.
E.g.
date,hour,...
10-1-2013,0,
10-1-2013,0,
10-1-2013,0,
10-1-2013,1,
10-1-2013,1,
How do I import using pandas so that that hour and date is combined or is that best done after the initial import?
df = DataFrame.from_csv('bingads.csv', sep=',')
If I do the initial import how do I combine the two as a date and then delete the hour?
Thanks
Define your own date_parser:
In [291]: from dateutil.parser import parse
In [292]: import datetime as dt
In [293]: def date_parser(x):
.....: date, hour = x.split(' ')
.....: return parse(date) + dt.timedelta(0, 3600*int(hour))
In [298]: pd.read_csv('test.csv', parse_dates=[[0,1]], date_parser=date_parser)
Out[298]:
date_hour a b c
0 2013-10-01 00:00:00 1 1 1
1 2013-10-01 00:00:00 2 2 2
2 2013-10-01 00:00:00 3 3 3
3 2013-10-01 01:00:00 4 4 4
4 2013-10-01 01:00:00 5 5 5
Apply read_csv instead of read_clipboard to handle your actual data:
>>> df = pd.read_clipboard(sep=',')
>>> df['date'] = pd.to_datetime(df.date) + pd.to_timedelta(df.hour, unit='D')/24
>>> del df['hour']
>>> df
date ...
0 2013-10-01 00:00:00 NaN
1 2013-10-01 00:00:00 NaN
2 2013-10-01 00:00:00 NaN
3 2013-10-01 01:00:00 NaN
4 2013-10-01 01:00:00 NaN
[5 rows x 2 columns]
Take a look at the parse_dates argument which pandas.read_csv accepts.
You can do something like:
df = pandas.read_csv('some.csv', parse_dates=True)
# in which case pandas will parse all columns where it finds dates
df = pandas.read_csv('some.csv', parse_dates=[i,j,k])
# in which case pandas will parse the i, j and kth columns for dates
Since you are only using the two columns from the cdv file and combining those into one, I would squeeze into a series of datetime objects like so:
import pandas as pd
from StringIO import StringIO
import datetime as dt
txt='''\
date,hour,A,B
10-1-2013,0,1,6
10-1-2013,0,2,7
10-1-2013,0,3,8
10-1-2013,1,4,9
10-1-2013,1,5,10'''
def date_parser(date, hour):
dates=[]
for ed, eh in zip(date, hour):
month, day, year=list(map(int, ed.split('-')))
hour=int(eh)
dates.append(dt.datetime(year, month, day, hour))
return dates
p=pd.read_csv(StringIO(txt), usecols=[0,1],
parse_dates=[[0,1]], date_parser=date_parser, squeeze=True)
print p
Prints:
0 2013-10-01 00:00:00
1 2013-10-01 00:00:00
2 2013-10-01 00:00:00
3 2013-10-01 01:00:00
4 2013-10-01 01:00:00
Name: date_hour, dtype: datetime64[ns]

Categories