Python Pandas - Return a new Timedelta with milliseconds floored resolution

To return a new Timedelta floored to this resolution, use the timedelta.floor() method. For milliseconds floored resolution, set the freq parameter to 'ms'.

Syntax

timedelta.floor(freq)

Parameters

freq: String representing the frequency. Use 'ms' for milliseconds resolution.

Example

Let's create a Timedelta object and floor it to milliseconds resolution ?

import pandas as pd

# Create a Timedelta object with nanoseconds precision
timedelta = pd.Timedelta('2 days 10 hours 45 min 20 s 35 ms 55 ns')

# Display the original Timedelta
print("Original Timedelta:")
print(timedelta)

# Floor to milliseconds resolution
floored_timedelta = timedelta.floor(freq='ms')

# Display the floored Timedelta
print("\nFloored to milliseconds:")
print(floored_timedelta)

Original Timedelta:
2 days 10:45:20.035000055

Floored to milliseconds:
2 days 10:45:20.035000

How It Works

The floor() method removes precision below the specified frequency. When using 'ms', it truncates nanoseconds, keeping only milliseconds precision. In our example, the original 55 nanoseconds are removed, leaving only the milliseconds component.

Other Frequency Options

import pandas as pd

timedelta = pd.Timedelta('2 days 10 hours 45 min 20 s 35 ms 55 ns')

print("Original:", timedelta)
print("Seconds floored:", timedelta.floor('s'))
print("Minutes floored:", timedelta.floor('min'))
print("Hours floored:", timedelta.floor('h'))

Original: 2 days 10:45:20.035000055
Seconds floored: 2 days 10:45:20
Minutes floored: 2 days 10:45:00
Hours floored: 2 days 10:00:00

Conclusion

The floor() method with freq='ms' truncates Timedelta precision to milliseconds. Use different frequency values like 's', 'min', or 'h' for other resolutions.