date arithmetic on strange date format

[Nico Peruzzi]

Hi Listers,

I have dates in the format YYYY-MM-DD HH:MM:SS.SSS and I need to find the
difference between 2 dates.

Unfortunately, the dates are strings, not dates, and I can't find that exact
format when I look at the date types available in SPSS.

2 options I see.

1. Anyone know how to convert this variable into a date format.  I could
then do the date arithmetic.
2. How about chopping off the HH:MM:SS.SSS (I don't need that level of
detail), and then I could probably throw it in Excel and do the date
arithmetic there.

Thanks in advance!  Nico

--
Nico Peruzzi, Ph.D.

[Sean McKenzie]

I receive dates like this 20060331  by various means i convert them to the
ussual 1/1/1900=1 convention.

A simple one is extract 1st 4 caharcter as text then 5-6 character then 7-8
and then concatenate

5-6/7-8/1-4 and convert to value dates...

A pain I know.  I am working with what I call a Miami Vice Type system, more
commonly known as a legacy system.

>From: Nico Peruzzi <[hidden email]>
>Reply-To: Nico Peruzzi <[hidden email]>
>To: [hidden email]
>Subject: date arithmetic on strange date format
>Date: Mon, 9 Oct 2006 14:50:46 -0700
>
>Hi Listers,
>
>I have dates in the format YYYY-MM-DD HH:MM:SS.SSS and I need to find the
>difference between 2 dates.
>
>Unfortunately, the dates are strings, not dates, and I can't find that
>exact
>format when I look at the date types available in SPSS.
>
>2 options I see.
>
>1. Anyone know how to convert this variable into a date format.  I could
>then do the date arithmetic.
>2. How about chopping off the HH:MM:SS.SSS (I don't need that level of
>detail), and then I could probably throw it in Excel and do the date
>arithmetic there.
>
>Thanks in advance!  Nico
>
>--
>Nico Peruzzi, Ph.D.

[Dennis Deck]

The best strategy is usually to parse the string dates into component
parts and construct date variables using functions (see COMPUTE syntax
and more complete documentation in Universals section). I usually nest
functions for compact code.

Where sDate1 and sDate2 are the input dates in string format:

COMPUTE Date1 = DATE.mdy(NUMBER(SUBSTR(sDate1,6,2),F2.0),
                         NUMBER(SUBSTR(sDate1,9,2),F2.0),
                         NUMBER(SUBSTR(sDate1,1,4),F4.0) .

COMPUTE Date2 = DATE.mdy(NUMBER(SUBSTR(sDate2,6,2),F2.0),
                         NUMBER(SUBSTR(sDate2,9,2),F2.0),
                         NUMBER(SUBSTR(sDate2,1,4),F4.0) .

COMPUTE Days = CTIME.days(Date2-Date1) .

FORMATS Date1 Date2 (ADATE10) / Days (F4.0).

Dennis Deck, PhD
RMC Research Corporation
[hidden email]

-----Original Message-----
From: Nico Peruzzi [mailto:[hidden email]]
Sent: Monday, October 09, 2006 2:51 PM
Subject: date arithmetic on strange date format

Hi Listers,

I have dates in the format YYYY-MM-DD HH:MM:SS.SSS and I need to find
the
difference between 2 dates.

Unfortunately, the dates are strings, not dates, and I can't find that
exact
format when I look at the date types available in SPSS.

2 options I see.

1. Anyone know how to convert this variable into a date format.  I could
then do the date arithmetic.
2. How about chopping off the HH:MM:SS.SSS (I don't need that level of
detail), and then I could probably throw it in Excel and do the date
arithmetic there.

Thanks in advance!  Nico

--
Nico Peruzzi, Ph.D.