These pages contain information regarding the output from a TRANSP run. The output files are written in a netCDF (Network Common Data Form) format, which is a set of software libraries and self-describing, machine-independent data formats that support the creation, access, and sharing of array-oriented scientific data. These are binary files and require a programming interface to be viewed and manipulated. The quickest way to visualize TRANSP output data is by using jetdsp, which enables a comparison with PPF data (measurements). Common data consistency checks performed with TRANSP are also described. The output signal names in the netCDF file are listed and described and a list of standard sets which read pre-defined lists of useful signals into jetdsp is given.
While jetdsp enables a fast comparison of the TRANSP output with measurements it is usually not the most ideal for data analysis. Python offers a great programming interface for netCDF files. Its use requires importing the netCDF4 library, which comes preinstalled on Heimdall pyhton installations, and is easy to install on private devices with scientific python packages like Anaconda. The three basic python commands to open a TRANSP output file and read its variables are:
from netCDF4 import Dataset # Import netCDF library
TRANSPDat = Dataset("/path_to_cdf/RunID.CDF") # Define the netCDF file path
TRANSP_NEUTT = np.array(TRANSPDat.variables["NEUTT"]) # Export individual variables as numpy arrays
There are two possible filel locations (path_to_cdf above) for TRANSP output storage on Heimdall - the first is the results directory where the output is stored immediately after the run is completed
and fetched, while the second is the warehouse directory where the output files are moved in case the run is archived:
Results dir.: /common/transp_shared/Data/result/JET/96482/V09/96482V09.CDF
Warehouse dir.: /common/transp_shared/Data/whouse/0096__/0096482/v09/0096482v09.cdf
Once you read in the desired variables, you treat these as standard numpy arrays - with TRANSP these will mainly be either 1D - time trace signals dependent on TRANSP
variable "TIME" - or 2D arrays - time evolved profiles dependent on TRANSP variables "TIME3" and "X"
(). OMFIT offers numerous built-in
plotting and analysis functions which you can use for constructing your own scripts. You can get more information through OMFIT's
video tutorials and by checking out the OMFIT script we use for plotting JET standard TRANSP output, available on
GitHub.
David Keeling has written an IDL widget to view TRANSP data. If you are unsold on JETdsp or are unfamiliar with python, you may wish to try this instead. Information can be found here. David has also written a script allowing TRANSP data to be read from the output netCDF file into an IDL session.
Information on how to prepare a TRAU PPF for reading TRANSP output into JETTO can be found in the appendix of the main google doc. Descriptions of the signals used can be found on the JETTO wiki.