DATE_TRUNC for rounding dates in PostgreSQL

Introduction

DATE_TRUNC helps you round a date or time to a specific part, like the start of the day or month. It makes dates simpler to compare or group.

When you want to find all records from the start of each month.

When you need to group sales by week, ignoring exact days.

When you want to compare events by hour, ignoring minutes and seconds.

When you want to reset a timestamp to the start of the day for easier filtering.

Syntax

PostgreSQL

DATE_TRUNC('part', timestamp)

The 'part' is the part of the date/time you want to keep, like 'day', 'month', or 'hour'.

The timestamp is the date or time value you want to round.

Examples

Rounds the timestamp to the start of the day (midnight).

PostgreSQL

SELECT DATE_TRUNC('day', TIMESTAMP '2024-06-15 14:23:45');

Rounds the timestamp to the first day of the month at midnight.

PostgreSQL

SELECT DATE_TRUNC('month', TIMESTAMP '2024-06-15 14:23:45');

Rounds the timestamp to the start of the hour (14:00:00).

PostgreSQL

SELECT DATE_TRUNC('hour', TIMESTAMP '2024-06-15 14:23:45');

Sample Program

This query shows how the same timestamp is rounded to the start of the day, month, and hour.

PostgreSQL

SELECT
  DATE_TRUNC('day', TIMESTAMP '2024-06-15 14:23:45') AS day_start,
  DATE_TRUNC('month', TIMESTAMP '2024-06-15 14:23:45') AS month_start,
  DATE_TRUNC('hour', TIMESTAMP '2024-06-15 14:23:45') AS hour_start;

Important Notes

DATE_TRUNC always returns a timestamp rounded down to the specified part.

You can use parts like 'second', 'minute', 'hour', 'day', 'week', 'month', 'year', and more.

It is very useful for grouping data by time periods in reports.

Summary

DATE_TRUNC rounds dates or times to a chosen part like day or month.

It helps simplify and group date/time data.

Use it when you want to ignore smaller parts of a timestamp.

Practice

1. What does the PostgreSQL function DATE_TRUNC('month', timestamp) do?

easy

A. It returns the exact timestamp without any change.

B. It rounds the timestamp down to the first day of the month at 00:00:00.

C. It rounds the timestamp up to the last day of the month at 23:59:59.

D. It converts the timestamp to a string showing only the month.

Solution

1. Step 1: Understand DATE_TRUNC behavior

   DATE_TRUNC cuts off smaller parts of a timestamp to the specified precision, rounding down.

2. Step 2: Apply 'month' precision

   Using 'month' sets the date to the first day of that month at midnight (00:00:00).

3. Final Answer:

   It rounds the timestamp down to the first day of the month at 00:00:00. -> Option B

4. Quick Check:

   DATE_TRUNC('month', timestamp) = first day of month [OK]

Hint: DATE_TRUNC cuts smaller parts, sets time to zero [OK]

Common Mistakes:

• Thinking it rounds up instead of down
• Assuming it returns a string
• Believing it keeps the original timestamp unchanged

2. Which of the following is the correct syntax to truncate a timestamp to the nearest hour in PostgreSQL?

easy

A. DATE_TRUNC('hour', 13:45:30)

B. DATE_TRUNC(hour, '2024-06-15 13:45:30')

C. DATE_TRUNC('hour', 2024-06-15 13:45:30)

D. DATE_TRUNC('hour', '2024-06-15 13:45:30')

Solution

1. Step 1: Check function syntax

   DATE_TRUNC requires the first argument as a string specifying the precision, e.g., 'hour'.

2. Step 2: Validate timestamp format

   The second argument must be a timestamp or string representing a timestamp enclosed in quotes.

3. Final Answer:

   DATE_TRUNC('hour', '2024-06-15 13:45:30') -> Option D

4. Quick Check:

   Correct syntax uses quotes around precision and timestamp [OK]

Hint: Always quote the precision string in DATE_TRUNC [OK]

Common Mistakes:

• Not quoting the precision string
• Passing timestamp without quotes
• Using incomplete timestamp formats

3. Given the query:
SELECT DATE_TRUNC('day', TIMESTAMP '2024-06-15 18:23:45');
What is the output?

medium

A. 2024-06-15 18:23:45

B. 2024-06-01 00:00:00

C. 2024-06-15 00:00:00

D. 2024-06-15 23:59:59

Solution

1. Step 1: Understand 'day' precision in DATE_TRUNC

   Truncating to 'day' sets the time to midnight (00:00:00) of the same date.

2. Step 2: Apply to given timestamp

   For '2024-06-15 18:23:45', truncation results in '2024-06-15 00:00:00'.

3. Final Answer:

   2024-06-15 00:00:00 -> Option C

4. Quick Check:

   DATE_TRUNC('day', timestamp) = date at midnight [OK]

Hint: Truncating to day resets time to 00:00:00 [OK]

Common Mistakes:

• Expecting original time to remain
• Confusing day truncation with month truncation
• Thinking it rounds to nearest day

4. Identify the error in this query:
SELECT DATE_TRUNC('min', '2024-06-15 12:34:56');

medium

A. The precision 'min' is invalid in DATE_TRUNC.

B. The timestamp string is not cast to TIMESTAMP type.

C. The function DATE_TRUNC requires three arguments.

D. The timestamp string should be in single quotes.

Solution

1. Step 1: Check precision argument

   DATE_TRUNC requires the first argument to be a specific precision string (e.g., 'minute', 'hour'). 'min' is invalid.

2. Step 2: Identify the error

   This causes an error like 'invalid value for parameter "field": "min"'.

3. Final Answer:

   The precision 'min' is invalid in DATE_TRUNC. -> Option A

4. Quick Check:

   Valid precisions include 'minute' [OK]

Hint: Use exact precision strings: 'minute' not 'min' [OK]

Common Mistakes:

• Assuming string is auto-converted
• Using invalid precision values
• Adding extra arguments to DATE_TRUNC

5. You want to group sales data by week starting on Monday using DATE_TRUNC. Which query correctly truncates the timestamp to the start of the week?

hard

A. SELECT DATE_TRUNC('week', sale_date + INTERVAL '1 day') - INTERVAL '1 day' FROM sales;

B. SELECT DATE_TRUNC('dow', sale_date) FROM sales;

C. SELECT DATE_TRUNC('month', sale_date) FROM sales;

D. SELECT DATE_TRUNC('week', sale_date) FROM sales;

Solution

1. Step 1: Understand DATE_TRUNC('week') behavior

   In PostgreSQL, DATE_TRUNC('week', timestamp) returns the start of the week on Sunday.

2. Step 2: Adjust to Monday start

   Adding 1 day before truncation shifts the week start to Monday, then subtracting 1 day corrects the date.

3. Final Answer:

   SELECT DATE_TRUNC('week', sale_date + INTERVAL '1 day') - INTERVAL '1 day' FROM sales; -> Option A

4. Quick Check:

   Shift date to align week start Monday [OK]

Hint: Shift date by 1 day to start week on Monday [OK]

Common Mistakes:

• Assuming 'week' starts on Monday by default
• Using invalid precision 'dow'
• Truncating to month instead of week