Difference between actual time and rawtime

[jg]

Hi,

There appears to be a difference between actual time and rawtime. This issue was flagged in 2003 (I believe) where it was stated that the difference was 1 hour and 1 minute. I have just done a calculation and it appears that the difference has grown to almost 2 hours.

Here are my results:

Consider the date July 8, 2009 8:00:00 EST.
Since the EST offset from UT (GMT) is 5:00, this date/time is equivalent to July 8, 2009 13:00:00 UT.
The Julian date for this is JD 2455021.04167

Consider the date January 1, 1970 0:00:00 EST.
Using the offset for UT, this is equivalent to January 1, 1970 05:00:00 UT. The Julian date for this is JD 2440587.70833

The difference between these two Julian dates is 14433.33334 days or 1,247,040,000.576 seconds

For the July 8, 2009 08:00 EST bar the rawtick function returns a time of 1,247,054,400 seconds

The difference between the Julian date difference and the rawtick return value is 14,399.424000024796 seconds which is just under 240 min (239.990 min).

Can someone please explain to me what I am missing in using the rawtick function. Perhaps the rawtick function does not account for leap seconds, etc. ... Does anyone know how rawtick is computed?

Any clarification of this matter would be greatly appreciated. Thank you in advance for any help you may be able to provide.

Regards,
Jane

[ACM]

Re: Difference between actual time and rawtime

Jane

Consider the date July 8, 2009 8:00:00 EST.
Since the EST offset from UT (GMT) is 5:00, this date/time is equivalent to July 8, 2009 13:00:00 UT.

On the date you indicate the timezone offset between ET and GMT is actually 4 and not 5 because of DST
Alex

Originally posted by jg
Hi,

There appears to be a difference between actual time and rawtime. This issue was flagged in 2003 (I believe) where it was stated that the difference was 1 hour and 1 minute. I have just done a calculation and it appears that the difference has grown to almost 2 hours.

Here are my results:

Consider the date July 8, 2009 8:00:00 EST.
Since the EST offset from UT (GMT) is 5:00, this date/time is equivalent to July 8, 2009 13:00:00 UT.
The Julian date for this is JD 2455021.04167

Consider the date January 1, 1970 0:00:00 EST.
Using the offset for UT, this is equivalent to January 1, 1970 05:00:00 UT. The Julian date for this is JD 2440587.70833

The difference between these two Julian dates is 14433.33334 days or 1,247,040,000.576 seconds

For the July 8, 2009 08:00 EST bar the rawtick function returns a time of 1,247,054,400 seconds

The difference between the Julian date difference and the rawtick return value is 14,399.424000024796 seconds which is just under 240 min (239.990 min).

Can someone please explain to me what I am missing in using the rawtick function. Perhaps the rawtick function does not account for leap seconds, etc. ... Does anyone know how rawtick is computed?

Any clarification of this matter would be greatly appreciated. Thank you in advance for any help you may be able to provide.

Regards,
Jane

[jg]

Re: Re: Difference between actual time and rawtime

Hi Alex,

Thank you for pointing out the error. You are absolutely correct. My apologies. I had forgotten about DST in the offset for July.

Here are my revised results:

Consider the date July 8, 2009 8:00:00 EST.
Since the EST offset from UT (GMT) is 4:00:00, this date/time is equivalent to July 8, 2009 12:00:00.0 UT.
The Julian date for this is JD 2455021.00000

Consider the date January 1, 1970 0:00:00 EST.
Using the offset of 5:00:00 for UT, this is equivalent to January 1, 1970 05:00:00 UT. The Julian date for this is JD 2440587.70833

The difference between these two Julian dates is 14433.29167 days or 1,247,036,400.288 seconds

For the July 8, 2009 08:00 EST bar the rawtick function returns a time of 1,247,054,400 seconds

The difference between the Julian date difference and the rawtick return value is 17,999.71199989319 seconds which is just under 300 min (299.995 min).

Alex, the time has drifted from the past post of 2003 regarding this issue.

Regards,
Jane

Originally posted by Alexis C. Montenegro
Jane



On the date you indicate the timezone offset between ET and GMT is actually 4 and not 5 because of DST
Alex

[ACM]

Re: Re: Re: Difference between actual time and rawtime

Jane

For the July 8, 2009 08:00 EST bar the rawtick function returns a time of 1,247,054,400 seconds

The difference between the Julian date difference and the rawtick return value is 17,999.71199989319 seconds which is just under 300 min (299.995 min).

I would expect that difference given the values used in your calculations. 300 minutes is in fact 5 hours
rawtime is the time elapsed from 1/1/1970 00:00:00 UTC. The rawtime value you are using however does not take into account the 5 hours of timezone offset which you are instead accounting for in your Julian date calculations
The rawtime at 05:00:00 UTC is 18,000 and not 0 (see screenshot enclosed below) so
1,247,054,400 - 18,000 = 1,247,036,400
This expression now replicates the calculations you are using in the Julian dates
So the difference between the Julian date calculations and rawtime is of 0.288 seconds and not 300 minutes
Alex

Originally posted by jg
Hi Alex,

Thank you for pointing out the error. You are absolutely correct. My apologies. I had forgotten about DST in the offset for July.

Here are my revised results:

Consider the date July 8, 2009 8:00:00 EST.
Since the EST offset from UT (GMT) is 4:00:00, this date/time is equivalent to July 8, 2009 12:00:00.0 UT.
The Julian date for this is JD 2455021.00000

Consider the date January 1, 1970 0:00:00 EST.
Using the offset of 5:00:00 for UT, this is equivalent to January 1, 1970 05:00:00 UT. The Julian date for this is JD 2440587.70833

The difference between these two Julian dates is 14433.29167 days or 1,247,036,400.288 seconds

For the July 8, 2009 08:00 EST bar the rawtick function returns a time of 1,247,054,400 seconds

The difference between the Julian date difference and the rawtick return value is 17,999.71199989319 seconds which is just under 300 min (299.995 min).

Alex, the time has drifted from the past post of 2003 regarding this issue.

Regards,
Jane

[jg]

Re: Re: Re: Re: Difference between actual time and rawtime

Hi Alex,

Thank you very much for your detailed explanation. I completely understand your explanation and it make sense. It however raises an an aspect of rawtime that I do not understand.

This result came about from using the following piece of code:

var firstBarIndex = getOldestBarIndex();
var myT = null;
if(null != firstBarIndex) {
myT = rawtime(firstBarIndex);
}

So it appears that this construction sends the bar time (at the bar index referenced) as local time and we then need to apply the timezone offset (accounting for daylight savings time, DST) to this value. Is there a cleaner way to do this? Perhaps by telling rawtime to convert to the correct UT time?

Thank you very much for all your help. I truly appreciate it.

Regards,
Jane

Originally posted by Alexis C. Montenegro
Jane



I would expect that difference given the values used in your calculations. 300 minutes is in fact 5 hours
rawtime is the time elapsed from 1/1/1970 00:00:00 UTC. The rawtime value you are using however does not take into account the 5 hours of timezone offset which you are instead accounting for in your Julian date calculations
The rawtime at 05:00:00 UTC is 18,000 and not 0 (see screenshot enclosed below) so
1,247,054,400 - 18,000 = 1,247,036,400
This expression now replicates the calculations you are using in the Julian dates
So the difference between the Julian date calculations and rawtime is of 0.288 seconds and not 300 minutes
Alex

[ACM]

Re: Re: Re: Re: Re: Difference between actual time and rawtime

Jane
As I understand it rawtime returns the bar time as UTC and not local time so you need to add the code to convert that to local time and not the other way round
Try searching the forum as this has been covered before and you should find several examples of how to handle this
Alex

Originally posted by jg
Hi Alex,

Thank you very much for your detailed explanation. I completely understand your explanation and it make sense. It however raises an an aspect of rawtime that I do not understand.

This result came about from using the following piece of code:

var firstBarIndex = getOldestBarIndex();
var myT = null;
if(null != firstBarIndex) {
myT = rawtime(firstBarIndex);
}

So it appears that this construction sends the bar time (at the bar index referenced) as local time and we then need to apply the timezone offset (accounting for daylight savings time, DST) to this value. Is there a cleaner way to do this? Perhaps by telling rawtime to convert to the correct UT time?

Thank you very much for all your help. I truly appreciate it.

Regards,
Jane

[jg]

Re: Re: Re: Re: Re: Re: Difference between actual time and rawtime

Hi Alex,

Thank you very much for all your help. I shall search the forums for the code suggested.

Regards,
Jane

Originally posted by Alexis C. Montenegro
Jane
As I understand it rawtime returns the bar time as UTC and not local time so you need to add the code to convert that to local time and not the other way round
Try searching the forum as this has been covered before and you should find several examples of how to handle this
Alex

[ACM]

Jane
FWIW regarding the difference of 0.288 secs the reason may lie in the rounding of the calculation from a date to JD and vice-versa. In fact using the USNO's converter I get the following when converting a date to JD
The Julian date for CE 1970 January 1 05:00:00.0 UT is
JD 2440587.70833
However when converting that same JD back to a date I then get the following
JD 2440587.70833 is
CE 1970 January 01 04:59:59.7 UT Thursday
which returns a difference of 0.3 secs
Alex

[jg]

Hi Alex,

Thank you very much for the follow up. I very much appreciate it.
In the interim I have discovered that the time discrepancy is due to roundoff (or truncation) in the last digit of the USNO's JD.

I coded USNO's Julian date converter and for 1970 January 1, 05:00:00.0 UT I get JD 2440587.7083333335. I also coded a converter to Gregorian calendar date and time, and for JD 2440587.7083333335 I get 1970/1/1 5h 0m 0s 0mS (milli seconds).

So the discrepancy in the rawtime is most likely due to truncation or roundoff in the Julian date.
Regards,
Jane

Originally posted by Alexis C. Montenegro
Jane
FWIW regarding the difference of 0.288 secs the reason may lie in the rounding of the calculation from a date to JD and vice-versa. In fact using the USNO's converter I get the following when converting a date to JD
The Julian date for CE 1970 January 1 05:00:00.0 UT is
JD 2440587.70833
However when converting that same JD back to a date I then get the following
JD 2440587.70833 is
CE 1970 January 01 04:59:59.7 UT Thursday
which returns a difference of 0.3 secs
Alex