แยกปี เดือน และวันในสัปดาห์

One datetime value packs a year, month, day, hour, and weekday inside it. The skill is pulling each piece out as its own column for analysis. 🧩

On a datetime column, the dt accessor is your gateway. It exposes every date part as a simple attribute you can read or assign.

Grabbing the year is one line. It is perfect for grouping sales or events into annual buckets.

The month attribute returns 1 through 12. Use it to compare the same month across different years.

Numbers are fine for sorting, but charts read better with names. Use month_name to get January, February, and friends.

The dayofweek attribute numbers days from Monday=0 to Sunday=6. It is the backbone of any weekday analysis.

Want readable labels instead? Use day_name to turn each date into Monday, Tuesday, and so on for clean grouping.

Combine dayofweek with a comparison to flag weekends. This boolean column instantly splits work days from rest days.

Two more handy parts: quarter for Q1 to Q4 reporting, and dayofyear for tracking position within a single year.

If your data has time, the dt accessor also exposes hour and minute. That makes it easy to study peak activity windows.

Each extracted part becomes a feature you can group by or feed to a model. This is how raw timestamps turn into insight.

You want to flag rows that fall on Saturday or Sunday.

You used the dt accessor to pull year, month, weekday, quarter, and hour from timestamps, turning each into a feature ready for grouping. 🗓️