CODEWITHSUNDEEP
pythonpandas
Jason
Jason

Reputation: 349

Creating time series data from pandas dataframe and known time interval

This data has no date time or other useful columns like other questions address.

I have a pandas data frame ( or single series) returned without date information like so : 

      [10:1]
0    2.264555
1    3.555655
2    4.023047
3    3.500272
4    3.033678
..        ...
152  3.808607
153  4.222035
154  5.205624
155  5.025041
156  5.711009

What I do have is the starting time and the time interval between each row.

Basically startingTime : 2020-02-19T02:54:29.000Z and timeInterval : "0.05" (5 seconds)

How could I go about creating an x,y style format like [[datetime, value], [datetime, value]] ?

Is this a situation where you would typically do the opposite of what i'm thinking and inject the computed date time as column in the data frame first?

Upvotes: 0

Views: 1162

Answers (2)

Pygirl
Pygirl

Reputation: 13349

You can create timestamp like this:

from datetime import datetime, timedelta
date_list = [str(datetime(2010, 2, 19, 2, 54, 29).strftime('%Y-%m-%d %H:%M:%S Z')) + str(timedelta(seconds=5*x)) for x in range(0, len(df))]

Result:

['2010-02-19 02:54:29 Z0:00:00',
 '2010-02-19 02:54:29 Z0:00:05',
 '2010-02-19 02:54:29 Z0:00:10',
 '2010-02-19 02:54:29 Z0:00:15',
 '2010-02-19 02:54:29 Z0:00:20',
 '2010-02-19 02:54:29 Z0:00:25',
.
.
.
]

then 

df.insert(0, 'TimeStamp', date_list)

Upvotes: 1

carnava1
carnava1

Reputation: 319

Creating a datetime column with fixed time interval:

import datetime

starting_time = datetime.datetime(2020,2,19,2,54,29)
td = 5  # timedelta
df['datetime'] = [starting_time + datetime.timedelta(seconds=i*td) for i in range(len(df))]

Then you can apply this to convert your values into list:

list(zip(df['datetime'], df['Column']))

Upvotes: 1

Related Questions