How to remove time part from a datetime in Kusto - azure-data-explorer

If I've got a Kusto datetime and I want to remove the time portion, leaving just a date at midnight, what's the best way? I can do this
todatetime(format_datetime( now(), "yyyy-MM-dd"))
but surely there's a more efficient way?

Use the startofday() function:
startofday( now() )
or the bin() function:
bin( now(), 1d )

Related

How do i subtract time from datetime in snowflake?

pdt.startTime is datetime
s_first.FromTimeOfDay is a time
I want to subtract the time drom the datetime. When i run the code below, Snowflake gives me this error invalid type [CAST(S_FIRST.FROMTIMEOFDAY AS TIMESTAMP_NTZ(9))] for parameter 'TO_TIMESTAMP_NTZ'
select (pdt.StartTime - (SELECT s_first.FromTimeOfDay::datetime FROM Shift s_first))
from RAW_CPMS_AAR.POWERBI_DowntimeTable AS PDT
When i try this:
select (pdt.StartTime::TIMESTAMP_NTZ(9) - (SELECT s_first.FromTimeOfDay::TIMESTAMP_NTZ(9) FROM Shift s_first))
from RAW_CPMS_AAR.POWERBI_DowntimeTable AS PDT
I get more or less the same error: invalid type [CAST(S_FIRST.FROMTIMEOFDAY AS TIMESTAMP_NTZ(9))] for parameter 'TO_TIMESTAMP_NTZ'
How do I convert the time into a datetime format so that I can subtract the two. It doesnt seem to me that there is a clear way to convert time into datetime in snowflake.
Is this what you're after?
select current_timestamp() as sample_timestamp
, time(sample_timestamp) as sample_time
, date(sample_timestamp) as sample_date;
A user pointed me in the right direction. i didnt realize i could use "dateadd" to also subtract time.
dateadd(HOUR, - (HOUR(current_timestamp())), temp.DateTime)

Converting Datetimes returns Null

I'm trying to convert a datetime that looks like this: 2017-09-19T07:00:00-07:00 into EST, but i keep getting Null values when using the hive built in UTC conversion.
I've tried using a regular expression to parse the date:
date_format(from_unixtime(unix_timestamp(2017-09-19T07:00:00-07:00, "yyyy-MM-dd'T'HH:mm:ss")
- (cast(regexp_extract(regexp_extract(2017-09-19T07:00:00-07:00, '(-[0-9][0-9]:[0-9][0-9])$', 1),'(-[0-9][0-9])',1) as int)*3600) -18000),'YYYY-MM-dd HH:mm')
but that's not good, since there's an hourly difference based on the time of year.
I've also tried:
FROM_UTC_TIMESTAMP(UNIX_TIMESTAMP(2017-09-19T07:00:00-07:00, "yyyy-MM-dd'T'hh:mm:ss:SSS'ZZZZZ'") * 1000, 'EST')
and
FROM_UTC_TIMESTAMP(UNIX_TIMESTAMP(2017-09-19T07:00:00-07:00, "yyyy-MM-dd'T'hh:mm:ss:SSS'Z'") * 1000, 'EST')
but that appears to not work either. What am I doing wrong?
I think that this method needs the date as a string like this:
date_format(from_unixtime(unix_timestamp('2017-09-19T07:00:00-07:00', "yyyy-MM-dd'T'HH:mm:ss")
Normally, the date formats are for strings, not for integers or numbers.
I found the answer on my own by combining the two ways of running the query.
date_format(
FROM_UTC_TIMESTAMP(
(unix_timestamp('2017-09-19T07:00:00-07:00', "yyyy-MM-dd'T'HH:mm:ss")
+ (cast(
regexp_extract(
regexp_extract('2017-09-19T07:00:00-07:00', '(-[0-9][0-9]:[0-9][0-9])$',1),'(-[0-9][0-9])',1) as int)
*-3600)
)*1000 ,'America/New York')
,'YYYY-MM-dd HH:mm:ss')
You are getting NULL because the pattern (format of the date and time) you have provided is not matching with the actual date time value. Correcting the date time format in your query would resolve this issue:
select from_unixtime(UNIX_TIMESTAMP("2017-09-19T07:00:00-07:00", "yyyy-MM-dd'T'HH:mm:ssXXX"), "yyyy-MM-dd HH:mm:ss");
Check out this link to know more about the date time patterns: https://docs.oracle.com/javase/7/docs/api/java/text/SimpleDateFormat.html

how to do sec_to_time in SQLite?

I have a field with seconds after midnight, and I have to convert it to HH:MM:SS. MySQL has sec_to_time(), which would be perfect, but SQLite does not.
How to convert?
I feel like it's some combination of strftime or date, and some dividing by 3600, but I can't get it.
For example, I need to convert 3601 to 1:01. Or, 32405 -> 9:05.
Use time()
select time(3601, 'unixepoch');
will give:
01:00:01

Calculate a date in behind 24 hours Hive

My demand is really so silly, so basically I need to go back in time 24 hours in a timestamp column in Hive.
So far, I have tried two different ways but it's not going thru:
select
recordDate, --original date
cast(date_sub(cast(recorddate as timestamp),1) as timestamp), -- going one day behind without hour
cast((cast(cast(recorddate as timestamp) AS bigint)-1*3600) as timestamp) -- crazy year
from mtmbuckets.servpro_agents_events limit 10;
My output looks:
I appreciate the support you can give me.
thanks
There is not straight forward function in hive .
1 Create UDF to do so .
or
Convert date in no of second and do you calculation( -24 *60*60) sec then change back int to data.
use from_unixtime and unix_timestamp to achieve below code.
select from_unixtime(unix_timestamp(recorddate) - 86400)
from mtmbuckets.servpro_agen ts_events limit 10;;
From_unixtime
Convert time string with given pattern to Unix time stamp (in seconds) The result of this function is in seconds.
Unix_timestamp
Converts time string in format yyyy-MM-dd HH:mm:ss to Unix timestamp (in seconds), using the default timezone and the default locale, return 0 if fail: unix_timestamp('2009-03-20 11:30:01') = 1237573801

SQLite3 DateTime comparison in linux

As per seen in image my first query return 5 rows but my second query does not return any rows.
It shoud be return 3 rows.
I also have tried with
Store my all datetime data in format of 'yyyy-MM-dd hh:mm:ss'
"SELECT billheaderid,billheadercode,billtotalitem,billtotalamount,createdby,createdon WHERE cretedon >= Datetime('2014-08-19 12:26:32')"
Date values with "AM/PM" fields cannot be compared correctly with string comparisons
(1 is larger than 0).
You have to change all the values in the database to the correct format yyyy-MM-dd hh:mm:ss.
(And it is not necessary to call the datetime function.)
Store your data in form of
'yyyy-MM-dd hh:mm:ss'
And please care that '2014-08-19 03:45 PM' must be store as '2014-08-19 15:45:23' not as '2014-08-19 03:45:23'.
After that you don't need use datetime function. I am sure it'll work 100%.

Resources