Hi,
How can we include a specific output variable in ResStock final parquet files report? Is that possible? Currently the variable is being reported for each building id in the results_timeseries.csv and was wondering if we can report the average value in the final parquet file?
Thanks,
Previous versions of ResStock let you specify EnergyPlus output variables to report beyond ResStock's default variables, using the PAT GUI. Examples include things like Zone Air Temperature for each thermal zone. It looks like newer versions of ResStock let you specify a more detailed output using the simulation_output_report key in the yml.
example
include_timeseries_component_loads: true
include_timeseries_zone_temperatures: true
However, it is possible modifying the simulation_output_report key won't write the specific EnergyPlus variables you are looking for to the output parquet file. You need to look at what the reporting measure (specifically ReportSimulationOutput) is doing to process E+ output variables into the ResStock output fields.
The ruby file for Report Simulation Output is located at restock\resources\hpxml-measures\ReportSimulationOutput
You can also add your own reporting measures as long as they "work within the ResStock workflow."
If you want to report an "average value," writing your own reporting measure could be your best bet. That, or post-processing your results in Excel or Pandas or whatever other tool you are familiar with.
Thank you both! @shorowit I am using the output_variables to report the variables of interest. But, they end up in the results_timeseries.csv file for each bldg. Is there a way to have them reported on the concatenated parquet files similar to some of the other timeseries?
Try iterating through results_timeseries.csv and creating your own concatenated parquet file? That's probably faster than digging through the resstock source code to figure out why output_variables aren't getting translated to the parquet.
Comments
Are you referring to the average of an individual building's timeseries or an average timeseries across multiple buildings or the averaging multiple building's annual energy use (this last one is the method my team has primarily used in our work to get a "representative" consumption value).
"Average value" is too vague.