Constructing the SourceData object

The first step is to construct a SourceData object. The SourceData specifies the location of the simulation results that should be used for computing the indicators, the years for which the indicators should be computed, and, optionally, a second data directory for which the indicators will be compared against. Each indicator requires a source data object to be passed to it. Every SourceData object accepts the following arguments:

• dataset_pool_configuration is an object that handles which datasets get loaded and in what order.

• cache_directory is a path to the directory containing the simulation results that the indicators should be computed from.

• comparison_cache_directory is an optional field. Set this field to the path to a second cache directory in order to run a cross-scenario indicator comparison. See 5.3.9 for more information.

• run_description is a description of this indicator batch. This field is optional.

• years are the default years that all the indicators that use this SourceData object will be computed for. This field is optional, although, if absent, every indicator will need to individually specify a years parameter to its constructor.

Here is an example. All indicators that use this SourceData object, and don't override the default, will be computed for 2010, 2015, and 2018.

source_data = SourceData(
   cache_directory = r'D:\urbansim_cache\run_1090.2006_11_14_12_12',
   comparison_cache_directory = r'D:\urbansim_cache\run_1091.2006_11_14_12_12',
   years = [2010, 2015, 2018],
   dataset_pool_configuration = DatasetPoolConfiguration(
         package_order=['urbansim','opus_core'],
         package_order_exceptions={},
         ),                  
)

The package_order parameter to dataset_pool_configuration needs to include all packages that the relevant indicators are defined for. For example, to include all indicators available for the Eugene application of UrbanSim, package_order=['eugene', 'urbansim', 'opus_core'].