Skip to content

Outputs: retrieving outputs with GemsPy

SimulationSession.run() returns a SimulationTable directly after solving:

from pathlib import Path
from gems_craft.study.folder import load_study
from gems_runner.session import SimulationSession
from gems_craft.optim_config import load_optim_config

study = load_study(Path("my_study"))
optim_config = load_optim_config(Path("my_study/input/optim-config.yml"))

session = SimulationSession(study=study, optim_config=optim_config)
results = session.run()  # SimulationTable

Via SimulationTableBuilder (low-level)

When using build_problem() directly, build the table from the solved problem:

from gems_runner.simulation.simulation_table import SimulationTableBuilder

results = SimulationTableBuilder().build(problem)

Accessing results

SimulationTable exposes a fluent accessor API and a raw DataFrame.

Fluent API

# All values for a component
component_view = results.component("gen_de")

# Pivot for a specific output variable
output_view = component_view.output("generation")

# Single value (single timestep, single scenario)
val = output_view.value(time_index=0, scenario_index=0)

# All timesteps for scenario 0
series = output_view.value(scenario_index=0)  # returns a pandas Series

Raw DataFrame

df = results.data

The DataFrame has columns: block, component, output, absolute_time_index, block_time_index, scenario_index, value, basis_status.

Reading the value of the optimisation variable var_id of component component_id for a single time step and scenario:

value = df[(df["component"] == component_id) & (df["output"] == var_id)]["value"].iloc[0]

For multi-time or multi-scenario results, filter additionally by block_time_index and scenario_index:

sub = df[(df["component"] == component_id) & (df["output"] == var_id)]
value_s0_t1 = sub[(sub["scenario_index"] == 0) & (sub["block_time_index"] == 1)]["value"].iloc[0]

Exporting results

results.to_csv(Path("output/"))       # writes one CSV per component
results.to_parquet(Path("output/"))   # writes Parquet files
results.to_netcdf(Path("output/"))    # writes a NetCDF file
ds = results.to_dataset()             # returns an xarray Dataset