Skip to content

Output directory and files

Nathan Moore edited this page Jul 6, 2020 · 9 revisions

Quick summary of the eagle output directory structure and the output files.

output_directory

Inside whatever folder is specified in output_directory in the yml file, the following folders and files exist.

  • dask_scheduler_##.json <---- dask details
  • dask_scheduler_##.out <--- dask details
  • job##.json <--- which simulations belong to which job
  • job.out-## <--- console/log output of buildstockbatch when running simulations contained in job ##
  • postprocessing.out <-- console/log output of postprocessing
  • sampling.out <--- console/log output of sampling
  • weather <--- weather files
  • housing_characteristics <--- the housing characteristics tsv files and sampled buildstock_csv files are here
  • results <--- simulation and postprocessing output folder

output_directory/results

  • parquet <--- postprocessed and combined timeseries parquet files
  • results_csvs <--- postprocessed results_csv files
  • simulation_output <--- raw simulation output for each buildings

output_directory/results/parquet

  • baseline <--- this is the equivalent of results_csvs but in parquet format
  • timeseries <--- the timeseries parquet files live here

output_directory/results/results_csvs

  • results_up##.csv.gz <--- results csv file for upgrade ##, gunzipped.

output_directory/results/simulation_output

  • results_job##.json.gz <--- merged results.json simulation output files for all simulation that ran in job ##
  • simulations_job##.tar.gz <--- collections of all the raw simulation output results (including the data_point.zip and singularity_output.log) for all buildings that were part of job ##
  • timeseries <--- folder containing the timeseries output for all buildings simulation, in parquet format