Let's check out those program debugging skills!

These are really great exercises, thanks for publishing them! I think they'll come in handy as training materials for analysts or reminders for us 'experienced' analysts.

It's great stuff and thanks for sharing it.

I have found what I think is a bug in the logic of the request (as well as data errors).

For example, storm 1962BERNICE has 3 records for 06SEP1962, 1 as a tropical storm and 2 as a Extratropical Cyclone...all with the same (and incorrect) max wind speed of 45 (should be 50, according to

https://en.wikipedia.org/wiki/1962_Pacific_hurricane_season#Tropical_Storm_Bernice).

Now, with the (fixed) logic of the program, it'll set MaxStatus as TS rather than EX I guess I'll claim that a cyclone is more than a tropical storm as so the max should be EX instead.

Just a fun thought...not in any way meant to take away from the great piece of work that it is.

Excellent find! Thanks for pointing that out!

The differences between determining the storms is a bit more complicated than only wind speed, but for the sake of the challenge I kept all the assumptions simple 😊.

I ran into the same discrepancies with the data and Wikipedia (same thing happens with category). One reason is the basic assumptions I made in the program. Another reason is most likely because Wikipedia has a 1-minute sustained peak intensity for wind speed, while the data has six-hourly information collected. I am not exactly sure how max wind speed is determined in that six-hour interval, but I think that's where that difference lies.

Another interesting storm to look at is StormKey "2014GENEVIEVE". Genevieve was the only storm missing a status description. According to Wikipedia, Genevieve was the first one to track across all three northern Pacific basins since Hurricane Dora in 1999.

https://en.wikipedia.org/wiki/Hurricane_Genevieve_(2014)#Intensification_and_peak_intensity

Looking at the details of Genevieve, you can see the status moves from HU to ST as the wind picks up. From my research (I am no expert on the subject) it seems that Genevieve may have been considered a super-typhoon (ST) since it crosses the Northwest Pacific at that time.

Difference between a hurricane and a typhoon

Only thing is that to be considered a super-typhoon you need 1-minute sustained surface winds of at least 150mph. Again, leaving us the question how is the max wind speed determined in that six-hour interval since it does not show 150+ in our data.

What is a super-typhoon?

Since the data we used is for Atlantic, Northeast and North Central Pacific, it causes the missing description for status!

I think understanding how the max wind speed is determined in that six hour interval would be great in answering our questions!

what I ended up doing in my project (which I call StormyDaniels.egp) is creating an alternative to the STORM_LOOKUP table:

data data.SaffirSimpson;
level=1; class='td'; desc='Tropical depression'; minspd=0; maxspd=38; output;
level=2; class='ts'; desc='Tropical storm'; minspd=39; maxspd=73; output;
level=3; class='c1'; desc='Category One'; minspd=74; maxspd=95; output;
level=4; class='c2'; desc='Category Two'; minspd=96; maxspd=110; output;
level=5; class='c3'; desc='Category Three'; minspd=111; maxspd=129; output;
level=6; class='c4'; desc='Category Four'; minspd=130; maxspd=156; output;
level=7; class='c5'; desc='Category Five'; minspd=157; maxspd=999; output;
run;

max(level) then gives me the description of any given storm. So, for Genevieve , I get 'Category Four' as max status.

I like the solution tomrvincent. Love that you adjusted the lookup table to your needs!

Debug 1

Input function couldn't Convert a numeric value to a character value using a format*

I just replaced

Category=input( MaxWindSpeed, StormCategory. );

with following

Category=put( MaxWindSpeed, StormCategory. );

and it woked.

Debug 2

Following

	if StormKey then call symputx('EndingYear',year(date));/*Create the last year macro variable*/

was replaced by

	if last.StormKey then call symputx('EndingYear',year(date));/*Create the last year macro variable*/

I thank you for this exercise. Very well done.

Glad you enjoyed it!