Question

The desire is for the user to instantiate a class that represents the transeint along with automatic access to a member item for each variable being represented (up to 200 variables). The set of variable class instances would be dynamic based on file based input data and the desire is to use the file provided variable names to create a collection of these variable instances that are accessible with a natural naming scheme. Effectively, the variable class hides the details of where the data is stored and the indepedent variable (ie, time) is stored. The following pseudo code expresses random lines that the end user may express. In some cases, the post processing may be much more extensive.

tran1 = CTransient('TranData', ...)
Padj = tran1.pressPipe1 + 10  # add 10 bar to a pressure for conservatism
Tsat = TsatRoutine( tran1.tempPipe1 )
MyPlotRoutine( tran1.tempPipe1, tran1.tempPipe2 )

where pressPipeX and tempPipeX names defined in the input data files and the corresponding numpy data vectors are specified in the ‘TranData’ file input file and are instances of a CVariable class.

Help on how to dynamically build the set of instances that represent the transient variables such that they can be accessed would be appreciated.

Answer 1

Your description of what you’re trying to do isn’t entirely clear, but automatically naming variables something1, something2, etc. are generally a bad idea. Use a list instead:

transientvariables = []
transientvariables.append(makenewtransientvariable())
# ...
for tv in transientvariables:
    print tv

Edit: OK, I think I see what you’re getting at, although your explanation still isn’t exactly easy to read. You have a collection of pipes, with a time series of temperature and pressure recorded for each one, right?

The easiest way would be to use a dictionary:

transients["tempPipe1"]

Or nested dictionaries:

transients["temp"]["Pipe1"]

Or you could override your class’ __getattr__ method, so that it looks in a dictionary, and you can do:

transients.tempPipe1

Edit 2: Overriding __getattr__ would look a bit like this:

def __getattr__(self, name):
    if name in self.varMap:
        return self.varMap[name]
    raise AttributeError