ST 445 -- Summer 2013

  Homework #1 -- Due THURSDAY, 23 May 2013

  In the file 'schrab.dat' in the 'sum13a' directory is an interesting 
  dataset from a doctoral student (Kevin Schrab) in meteorology who 
  studied tornados. Each record corresponds to a storm cell.  For each 
  storm cell in an outbreak, we have CELLID (cols 1-4), UMAX (outflow)
  (cols 11-12), MDA (max deviation angle)(cols 14-16), FMAX (see below)
  (col 18), SFCVOR (surface vorticity)(cols 20-21), PBE (potential 
  buoyant energy)(cols 23-26), SRAW (storm-relative ambient wind)(cols 
  28-31), and outbreak number (cols 33-34).  

  (a) Write a SAS program to read in this information, skipping the 
  variable in columns 6-9. Label each variable as appropriate. 

  (b) The first two characters of the CELLID are state postal codes 
  used to name outbreaks.  Create a variable (as part of the input 
  statement) with just those two characters.  

  (c) Tornado intensity is measured on the Fujita scale that goes from 
  F0 (weak) to F5 (strong).  In this file, FMAX measures intensity, 
  where zero denotes no tornado, and, if a tornado occurred, one was 
  added to the Fujita scale value.  Recode with a new intensity variable 
  so that no tornado corresponds to a missing value and the correct 
  Fujita scale value is given (that is, subtract one).

  (d) Create another variable, this time a character variable, that 
  takes the value 'strong' when FMAX is 5 or 6 (corresponding to F4 or 
  F5), 'no' when there is no tornado, and 'weak' for the other cases.

  (e) (*optional*) Print out the first few observations.
  
  (f) Using SAS and the variable created in (d), find the means of UMAX
  and MDA when the cell produced no, weak, or strong tornados.

  (g) Using SAS, construct a table with state code, number of cells, 
  number of cells that produced tornados, and maximum intensity, with 
  as little other information as possible.  Use titles and variable 
  labels as appropriate. 

  (h) Plot intensity against UMAX. (A simple proc plot is ok)

  (i) (*optional* and not easy) Use SAS to determine whether the 
  outbreak numbers are unique to the state postal codes (and vice 
  versa).

  FOR ALL EXERCISES: Hand in (paper) 
  (1) The answers to any questions asked. (none this time)
  (2) Your program * With at least as many comments as you have toes! * 
  (3) Your output