I have a dataframe
with meteorological
data every 30 minutes
. With my datetime index I need to create a column with timestamps
, but it must be in decimal
. Here's the example below:
In [134]: df.index[0:3]
Out[134]:
DatetimeIndex(['2016-01-01 00:30:00', '2016-01-01 01:00:00',
'2016-01-01 01:30:00'],
dtype='datetime64[ns]', name='date_time', freq=None)
I need create a column as follows:
df.new[0:3]
0.5,1,1.5
Where have 30 minutes i transform in .5
.
Follow my script:
import pandas as pd
import numpy as np
df = pd.read_csv('./cs_teste_full_output_2018-02-26T004329_adv.csv',skiprows=(0),
header=1,na_values='-9999.0')
df = df.drop(df.index[[0]])
df['date_time'] = df['date'] + str(' ') + df['time']
df = df.set_index(pd.DatetimeIndex(df['date_time']))
df.index.strftime('%M')/60
for i in range(1,len(df.index),1):
print(i)
df['minute'][i] = np.array(list(map(int,list(df.index.strftime('%M')))))/60
df['hour'] = df.index.strftime('%H')
df['hour_minute'] = df['hour'] + df['minute']
But this way it is not working and I can not do it any other way.
One way is to extract the hour and convert minutes to hours.
There should be no need to convert to / from strings.
import pandas as pd
idx = pd.DatetimeIndex(['2016-01-01 00:30:00',
'2016-01-01 01:00:00',
'2016-01-01 01:30:00'],
dtype='datetime64[ns]', name='date_time', freq=None)
idx.hour + idx.minute / 60
# Float64Index([0.5, 1.0, 1.5], dtype='float64', name='date_time')
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With