Two long but useful days at Melbourne Uni have just wound down, and five dozen meteorologists and climate scientists have given us feedback on a well-run bootcamp:
Good
Bad
splitting big problems into small pieces
SQLite is easy
functions instead of loops
Python was easy
naming functions after pets
hands-on application at the same time as instructor
morning tea
dictionaries
explaining why, rather than just what
swapping variables
using our own computers
copy/paste onto Etherpad
arrays in NumPy
sticky notes
a bit of motivational speaking
location (I'm not from Melbourne, eh)
Subversion FTW!
Git + BitBucket
Python Notebook
hey, there's useful stuff online
helpers
red/green/refactor
really practical exampels of test-driven development
Greg tells stories
lot of participation
talk to other people during the breaks
regular expressions
cheap!
seeing how programmers think
realizing how much room for growth I have
scared in the first ten minutes...
the material is free
peer instruction
WiFi
didn't do any meteorological data sets
the jokes
viewing angle
installing stuff
couldn't use some advanced stuff (I'm a beginner)
coffee
running right after conference
squeezing four half days into two full days (maybe try 3?)
not enough exercises
not long enough
content could have been more focused
didn't do NetCDF files
a bit of motivational speaking
didn't see comparison of Python with R
lot of spread of material/ability
longer breaks (I'm so tired)
databases irrelevant to me
IPython/NumPy confusing
wasn't in New Zealand
not enough time on version control
wasn't in Fiji
hated SQL
Greg speaks waaaaaaaaay too fast
learned a lot about Vi
hopping
pretty intense, dude
took a long time to restart after breaks
more on plotting
lighting in the room is uneven
more about linking Python to other things
...after first ten minutes, it was OK
Originally posted 2013-02-15 by Greg Wilson in University of Melbourne.