NumPy: Convert numpy datetime64 to Timestamp
Write a NumPy program to convert numpy datetime64 to Timestamp.
Sample Solution:-
Python Code:
# Importing the required libraries
import numpy as np
from datetime import datetime
# Getting the current date and time in UTC
dt = datetime.utcnow()
# Displaying the current date and time
print("Current date:")
print(dt)
# Converting the datetime object to a NumPy datetime64 object
dt64 = np.datetime64(dt)
# Calculating the timestamp from the datetime64 object
# The timestamp is calculated by subtracting the datetime '1970-01-01T00:00:00Z' from the current datetime
# Then it's divided by the number of seconds in one unit of timedelta64 to get the timestamp
ts = (dt64 - np.datetime64('1970-01-01T00:00:00Z')) / np.timedelta64(1, 's')
# Displaying the timestamp
print("Timestamp:")
print(ts)
# Converting the timestamp back to UTC datetime and displaying it
print("UTC from Timestamp:")
print(datetime.utcfromtimestamp(ts))
Sample Output:
Current date: 2017-04-01 08:01:12.722055 Timestamp: 1491033672.72 UTC from Timestamp: 2017-04-01 08:01:12.722055
Explanation:
In the above exercise –
dt = datetime.utcnow() - Get the current date and time using the datetime.utcnow() function and store it in the dt variable.
dt64 = np.datetime64(dt) - Convert the dt variable to a numpy datetime64 object using the np.datetime64() function and store it in the dt64 variable.
ts = (dt64 - np.datetime64('1970-01-01T00:00:00Z')) / np.timedelta64(1, 's') - Calculate the number of seconds that have elapsed since the Unix epoch (January 1, 1970 at 00:00:00 UTC) by subtracting the Unix epoch datetime from dt64, dividing the result by a numpy timedelta64 object representing one second, and storing the result in the ts variable.
print(datetime.utcfromtimestamp(ts)) - Convert the ts variable to a human-readable datetime object in UTC timezone using the datetime.utcfromtimestamp() function and finally print the result.
Pictorial Presentation:
Python-Numpy Code Editor: