Getting data out of CLIDE¶

Nicolas Fauchereau has written several helper functions in python, with the aim of mimicking the functions that are available in R from sourcing the common/clidesc.r file

These functions are contained in the file clidesc/common/clide.py which (for now and until we create a conda package and host it on binstar) needs to be copied over to the same common directory.

At the beginning of each Python script, one must import these functions (the equivalent of calling source() in a R script)

This is done by entering these lines:

import sys, os

Local = False # set to True for testing purposes

if local:
    sys.path.append('../common')
    from clide import *
else:
    source_path = os.path.dirname(os.path.realpath(sys.argv[0]))
    sys.path.append(os.path.join(source_path, '../common'))
    from clide import *
    conn = clide_open(base_path)

which will have for effect to i) make all the functions implemented in clide.py available for calling from within the python script, ii) parse the command line arguments and iii) establish the connection to the CLIDE database (which will be contained in the conn object).

These functions, their call sign and outputs are briefly described below: note that when returning a table (e.g. the result of a SQL query to the clide database) this table is a Pandas DataFrame object, which is similar to (but more efficient than!) a R dataframe, and makes slicing, columns or row selection, resampling etc extremely convenient. When time-series are returned, the index of the dataframe (i.e. the ‘rows’ identifier) is a Pandas datetime index object resulting from the conversion of the lsd field to a python datetime object and setting it as the index for the DataFrame. The name of the index is invariably timestamp.