I found a weird thing. If a timestamp value subtract another, then Redshift will return an strange prefix. For example,
select table1.c_timestamp - table1.c_timestamp from table_1
Expect result should be ZERO or similar something, because these two timestamp values are same.
However, what I received is "5012369 years 4 mons", which I have no idea how does Redshift calculate the result.
Is there anyone can show me some clues?
Thanks
Contrary to the other answer,
Datediff doesn't exactly subtract, but rather counts the number of times the datepart chosen starts between the two timestamps.
datediff(second, '2018-04-10 00:00:00.001','2018-04-10 00:00:00.999')
>> 0
select datediff(second, '2018-04-10 00:00:00.999','2018-04-10 00:00:01.001')
>> 1
See: Datediff documentation
Edit: this is the way I found of how to perform the OP's task
SELECT
round(((EXTRACT('epoch' FROM TIMESTAMP '2018-05-27 09:59:59.999') - EXTRACT('epoch' FROM TIMESTAMP '2018-05-27 09:59:59.001'))*1000 + EXTRACT(millisecond FROM TIMESTAMP '2018-05-27 09:59:59.999') - EXTRACT(millisecond FROM TIMESTAMP '2018-05-27 09:59:59.001'))::real/1000)
The right way to subtract between datetimes is:
select datediff(seconds, table1.c_timestamp, table1.c_timestamp) from table_1
Of course, it doesn't make much sense to subtract a timestamp from itself, because that obviously returns 0, but I assume you just run that as a test.
Related
I want to get truncked data over the last month. My time is in unix timestamps and I need to get data from last 30 days for each specific day.
The data is in the following form:
{
"id":"648637",
"exchange_name":"BYBIT",
"exchange_icon_url":"https://cryptowisdom.com.au/wp-content/uploads/2022/01/Bybit-colored-logo.png",
"trade_time":"1675262081986",
"price_in_quote_asset":23057.5,
"price_in_usd":1,
"trade_value":60180.075,
"base_asset_icon":"https://assets.coingecko.com/coins/images/1/large/bitcoin.png?1547033579",
"qty":2.61,
"quoteqty":60180.075,
"is_buyer_maker":true,
"pair":"BTCUSDT",
"base_asset_trade":"BTC",
"quote_asset_trade":"USDT"
}
I need to truncate data based on trade_time
How do I write the query?
The secret sauce is the date_trunc function, which takes a timestamp with time zone and truncates it to a specific precision (hour, day, week, etc). You can then group based on this value.
In your case we need to convert these unix timestamps javascript style timestamps to timestamp with time zone first, which we can do with to_timestamp, but it's still a fairly simple query.
SELECT
date_trunc('day', to_timestamp(trade_time / 1000.0)),
COUNT(1)
FROM pings_raw
GROUP BY date_trunc('day', to_timestamp(trade_time / 1000.0))
Another approach would be to leave everything as numbers, which might be marginally faster, though I find it less readable
SELECT
(trade_time/(1000*60*60*24))::int * (1000*60*60*24),
COUNT(1)
FROM pings_raw
GROUP BY (trade_time/(1000*60*60*24))::int
I have a column in my dataset that has a datatype of bigint:
Col1 Col2
1 1519778444938790
2 1520563808877450
3 1519880608427160
4 1520319586578960
5 1519999133096120
How do I convert Col2 to the following format:
year-month-day hr:mm:ss
I am not sure what format my current column is in but I know that it is supposed to be a timestamp.
Any help will be great, thanks!
Have you tried to use functions like from_unixtime? You could use it to convert unix time to timestamp, then you could use date_format to display it in way you want. Notice that in your example your unix time is with microseconds, so you might want to convert it first to milliseconds.
I have not tested that but I am assuming that your code should look like:
date_format(from_unixtime(col2/1000), '%Y-%m-%d %h:%i:%s')
Notice that from_unixtime accepts also a time zone.
Please visit this page to see the more details about date related functions: https://docs.starburstdata.com/latest/functions/datetime.html
I believe the denominator should be 1000000 not 1000. Probably a typo. Anyways juts adding the test results here for others reference.
-- Microseconds
select date_format(from_unixtime(cast('1519778444938790' as bigint)/1000000), '%Y-%m-%d %h:%i:%s');
2018-02-28 12:40:44
If you need to filter the data where the column is in BIGINT Unix format, then you can use the following snippet to compare : from_unixtime(d.started_on /1000) >= CAST('2022-05-10 22:00:00' AS TIMESTAMP )
Accepted answer is a bit misleading. You should divide by 1000.0 otherwise you'll lose ms precision and be limited to second precision:
date_format(from_unixtime(col2/1000.0), '%Y-%m-%d %h:%i:%s')
I have two timestamps and I would like to have a result with the difference between them. I found a similar question asked here but I have noticed that:
select
to_char(column1::timestamp - column2::timestamp, 'HH:MS:SS')
from
table
Gives me an incorrect return if these timestamps cross multiple days. I know that I can use EPOCH to work out the number of hours/days/minutes/seconds etc but my use case requires the result as a timestamp (or a string...anything not an interval!).
In the case of multiple days I would like to continue counting the hours, even if it should go past 24. This would allow results like:
36:55:01
I'd use the built-in date_part function (as previously described in an older thread: How to convert an interval like "1 day 01:30:00" into "25:30:00"?) but finally cast the result to the type you desire:
SELECT
from_date,
to_date,
to_date - from_date as date_diff_interval,
(date_part('epoch', to_date - from_date) * INTERVAL '1 second')::text as date_diff_text
from (
(select
'2018-01-01 04:03:06'::timestamp as from_date,
'2018-01-02 16:58:07'::timestamp as to_date)
) as dates;
This results in the following:
I'm currently unaware of any way to convert this interval into a timestamp and also not sure whether there is a use for it. You're still dealing with an interval and you'd need a point of reference in time to transform that interval into an actual timestamp.
I have found similar help, but the issue was more complex, I am only good with the basics of SQL and am striking out here. I get a handful of columns a,b,c,startdate,enddate and i need to parse that data out into multiple rows depending on how many months are within the range.
Eg: a,b,c,1/1/2015, 3/15,2015 would become:
a,b,c,1/1/2015,value_here_doesnt_matter
a,b,c,2/1/2015,value_here_doesnt_matter
a,b,c,3/1/2015,value_here_doesnt_matter
Does not matter if the start date or end date is on a specific day, the only thing that matters is month and year. So if the range included any day in a given month, I'd want to output start days for each month in the range, with the 1st as a default day.
Could I have any advice on which direction to begin? I'm attempting generate_series, but am unsure if this is the right approach or how to make it work with keeping the data in the first few arbitrary columns consistent.
I think generate_series is the way to go. Without knowing what the rest of your data looks like, I would start with something like this:
select
a, b, c, generate_series(startdate, enddate, interval '1 month')::date
from
my_table
I am converting an Unix script with a SQL transact command to a PostgreSQL command.
I have a table with records that have a field last_update_time(xtime) and I want to select every record in the table that has been updated within a selected period.
Say, the current time it 05/01/2012 10:00:00 and the selected time is 04/01/2012 23:55:00. How do I select all the records from a table that have been updated between these dates. I have converted the 2 times to seconds in the Unix script prior to issuing the psql command, and have calculated the interval in seconds between the 2 periods.
I thought something like
SELECT A,B,C FROM table
WHERE xtime BETWEEN now() - interval '$selectedtimeParm(in secs)' AND now();
I am having trouble evaluating the Parm for the selectedtimeParm - it doesn't resolve properly.
Editor's note: I did not change the inaccurate use of the terms period, time frame, time and date for the datetime type timestamp because I discuss that in my answer.
What's wrong with:
SELECT a,b,c
FROM table
WHERE xtime BETWEEN '2012-04-01 23:55:00'::timestamp
AND now()::timestamp;
If you want to operate with a count of seconds as interval:
...
WHERE xtime BETWEEN now()::timestamp - (interval '1s') * $selectedtimeParm
AND now()::timestamp;
Note the standard ISO 8601 date format YYYY-MM-DD h24:mi:ss which is unambiguous with any locale or DateStyle setting.
The first value for the BETWEEN construct must be the smaller one. If you don't know which value is smaller use BETWEEN SYMMETRIC instead.
In your question you refer to the datetime type timestamp as "date", "time" and "period". In the title you used the term "time frames", which I changed to "timestamps". All of these terms are wrong. Freely interchanging them makes the question even harder to understand.
That, and the fact that you only tagged the question psql (the problem hardly concerns the command line terminal) might help to explain why nobody answered for days. Normally, it's a matter of minutes around here. I had a hard time understanding your question.
Understand the data types date, interval, time and timestamp - with or without time zone. Start by reading the chapter "Date/Time Types" in the manual.
Error message would have gone a long way, too.
For anyone who is looking for the fix to this. You need to remove timestamp from the where clause and use BETWEEN!
TABLENAME.COL-NAME-FOR-TIMESTAMP BETWEEN '2020-01-29 04:18:00-06' AND CURRENT_TIMESTAMP