|
A trip to the library
|
NCI’s public data catalogue is at https://datacatalogue.nci.org.au
Additional datasets are available from ARCCSS CMS http://climate-cms.unsw.wikispaces.net/Data
THREDDS allows you to selectively inspect and download data
NetCDF format combines data and meta-data such as units and history
The Siphon library can scan THREDDS indexes like a directory using Python
|
|
Using NetCDF in Python
|
Xarray lets you read NetCDF files in Python and inspect their contents
NetCDF fields are made up of data + dimensions + metadata
Numpy and Scipy let you perform calculations on fields
|
|
Optimising analyses
|
The easiest way to be faster is to do less work
Reading the opposite end of an array is slow, reading from disk is slower
Filling up memory makes the computer grind to a halt
|
|
Outside Python
|
|
|
Publishing Data
|
|
|
Analysing model output
|
|