Working with Date and

1
Chapter 4

• Working with Date and
• Longitudinal Data

2
_at_

• data example
• input _at_4 x _at_8 y
• datalines
• 11 134 556 889
• 15 221 441 449
• 21 101 552 331

• The _at_ tells SAS which column to start reading the
  data.
• We dont have to read in all 4 variables, only
  the ones we want.

3
Informats

• Informats are of the form W.n.
• W is how many columns there are.
• n tells how many digits there are to the right of
  the decimal
• For example,
• The informat for 44221.8 is 6.1. Thats 6
  columns for all digits with 1 decimal point.
• The informat for 54422.18 is 7.2.

4
Working with dates

• SAS is able to read in dates in many formats.
• For example, SAS can read in the date
• September 11, 2001 in any of the following
• ways.
• 091101
• 09112001

5

• 09/11/01
• 09/11/2001
• 110901
• 11092001
• 010911
• 11Sep2001
• And many others.

• See page 124 in the text to see the list of all
  date formats.
• We only need to tell SAS what format the date is
  in.

6
Date Example

• data patients
• input _at_1 dob MMDDYY8. _at_10 kind MMDDYY10.
• _at_21 hschool MMDDYY6.
• ageinkindkind-dob
• ageinhshschool-dob
• 04/11/80 08/22/1985 082493
• 06/13/88 08/21/1993 082301
• 04/21/80 08/20/1985 082193

7

• Once SAS knows it is reading a date, it
• converts the date into the number of days from
• January 1, 1960.
• Dec 31, 1959 is -1
• Jan 2, 1960 is 1

8

• The problem is this Suppose someones
• birthdate is 08/22/01.
• Will SAS understand this to mean 8/22/1901
• or 8/22/2001?

9

• If you use options yearcutoff1900 then SAS
• will recognize that the relevant years are from
• 1900 to 1999 in which case this date means
• 8/22/1901.
• This cutoff marks the beginning of a 100-year
• interval. Starting in SAS version 7, the
  default
• cutoff is 1920.

10
Longitudinal Data

• This is data where repeated measurements are
• taken over time.
• Examples
• Recording a patients heart rate, blood pressure,
  doctors fee each time they visit.
• Recording a trees height each month.
• Recording the temperature each day.

11
Patients Data Set

• There are 5 patients. Their gender, time in the
• doctors office and weight is recorded at each
• visit.
• We may want
• a listing of each patients weight at the time of
  their last visit.
• the difference in their weight from the first
  visit to the last visit

12

• To determine the persons weight at the time of
• their last visit, read the data in as usual.
• Then
• proc sort datapatients
• by id
• data lastwt
• set patients
• by id
• if last.id
• run