Output files
When AMES is run, it generates several output files in the following folders:
otuputs
└───NAME
└───calibration
└───[diagnostics]
└───results
The NAME
of the folder under outputs
is specified by the output_folder
parameter in the configuration file. The diagnostics
folder is only present if the report-diagnostics
parameter is set to true
in the configuration file.
Different scenarios can be specified by different configuration files, with different output_folder
parameters set. In that case, there will be several folders:
otuputs
└───NAME 1
└───calibration
└───[diagnostics]
└───results
└───NAME 2
└───calibration
└───[diagnostics]
└───results
...
In addition, AMES writes a log file called AMES_log_XXXX.txt
to the main folder, where XXXX
is the output_folder
specified in the configuration file. If the model fails, then the error message will be written to the log file.
The AMES model is meant to be run together with LEAP, as explained in the page on the LEAP-AMES link. LEAP calculates demand for energy and energy production, while AMES covers the rest of the economy – the energy sectors specified in the configuration file are excluded from AMES's calculations. However, for calibration it can be useful to run AMES separately from LEAP, with the energy sectors included. This is done by setting include_energy_sectors = true
when calling the AMES.run()
function (see Running the AMES model for details.)
When AMES is run with energy sectors included, the output folder is named XXXX_full
, where XXXX
is the output_folder
specified in the configuration file.
The diagnostics
folder
When the report-diagnostics
parameter is set to true
, a variety of diagnostic outputs are reported.
One crucial diagnostic indicator, which is motivated in Isolating the energy sector is given in the file nonenergy_energy_link_measure.txt
. It contains a message like the following:
Measure of the significance to the economy of the supply of non-energy goods and services to the energy sector:
This value should be small: 2.52%.
Whether the estimated parameter is sufficiently small or not depends on the purposes of the analysis. For the Freedonia sample model, it suggests that excluding the demand by the energy sector of non-energy goods and services could lead to a 2.52% discrepancy.
An additional set of files with names such as model_0_2010.txt
provide an export of the linear goal program prepared by the AMES model. Examining these files can sometimes be helpful when the log file indicates an error in JuMP (the Julia mathematical programming library). The file offers an explicit formulation of the model, with the calculated parameters:
Min 1.7333333333333334 ugap[1] + 0.5333333333333333 ugap[2] + 0.6000000000000001 ugap[3] + ...
Subject to
eq_util[1] : ugap[1] + u[1] == 1.0
eq_util[2] : ugap[2] + u[2] == 1.0
eq_util[3] : ugap[3] + u[3] == 1.0
eq_util[4] : ugap[4] + u[4] == 1.0
...
Some of the files provide "sanity checks" on the input data.
demand_coefficients.csv
: These should sum to a total less than one down the columns[energy_share.csv]
: If present, these should be less than one (only reported if energy sectors are excluded)imported_fraction.csv
: These should be less than oneprofit_margins.csv
: These should be greater than 1 but (usually) less than 2supply_fractions.csv
: These should sum to one along columns (unless the product is not produced domestically, in which case the sum will be zero)
Other files report figures that should be close to, although not necessarily identical to, the corresponding values in the supply-use table. The differences arise because AMES compensates for territorial adjustments or stock changes:
domestic_production.csv
exports.csv
final_demand.csv
imports.csv
investment.csv
margins.csv
sector_output.csv
tot_intermediate_demand_all_products.csv
tot_intermediate_supply_all_sectors.csv
tot_intermediate_supply_non-energy_sectors.csv
: Note that these values will be lower than in the supply-use table due to the excluded energy sectorswages.csv
If intermediate demand coefficients are calculated endogenously, then a further set of files will be reported containing tables of intermediate demand coefficients for each year of the simulation:
demand_coefficients_[year].csv
The calibration
folder
The files in this folder are generated after further adjustments and a calibration run of the linear goal program. Some of these files can also be used as a "sanity check", including:
basic_prices_0.csv
: These should be close to 1capacity_utilization_0.csv
: These should be less than or equal to 1, and most should be close to 1capital_output_ratio_0.csv
: These should normally lie between 1 and 4, but may be outside that rangemargins_neg_0.csv
andmargins_pos_0.csv
: These should sum to the same totalwage_share_0.csv
: These should be less than 1
Others of the files in the calibration
folder have names similar to the diagnostics
folder. The values may be noticeably different due to the adjustments carried out by the AMES model:
domestic_production_0.csv
exports_0.csv
final_demand_0.csv
imports_0.csv
sector_output_0.csv
tot_intermediate_supply_non-energy_sectors_0.csv
The results
folder
The results
folder contains the simulation results from the model. Once the model is fully developed, calibrated, and running in practice, this is the most interesting folder to look at.
When running with LEAP, the AMES model may be run several times to converge on a consistent set of results. The results folder contains outputs from each of the runs:
By default, the AMES model reports results only for non-energy sectors. To include all sectors, set the option include_energy_sectors = true
when running the AMES model.
Two of the results files (with the run number indicated by #
) contain multiple variables:
indices_#.csv
: The indices that are passed to LEAP, as specified in the configuration file (if no indices are defined, then this file will not be written)collected_variables_#.csv
: A set of key variables, such as net exports, GDP, and so on
AMES is a demand-driven model, so economic growth is driven by final demand, exports net of imports, and investment. Total investment expenditure is reported in collected_variables_#.csv
. Final demand, exports, and imports are reported by product:
final_demand_#.csv
exports_#.csv
andimports_#.csv
Output by sector is reported as actual output and potential output (that is, output at full capacity utilization). The ratio of actual to potential output is capacity utilization, which is reported separately:
sector_output_#.csv
potential_sector_output_#.csv
capacity_utilization_#.csv
Real value added by sector – that is, sector output less the cost of intermediate goods and services – is also reported:
real_value_added_#.csv
A further sector indicator, domestic insertion, is a measure of direct and indirect demand for domestic products by sector:
domestic_insertion.csv
Investment rates are determined by the autonomous investment rate, the profit rate, capacity utilization, and the interest rate. The autonomous investment rate is an expectation of future growth, which gradually adjusts over time in response to realized growth (that is, the model features "adaptive expectations"). The interest rate is reported in collected_variables_#.csv
, while the capacity utilization file was listed above. The other relevant files are:
profit_rate_#.csv
autonomous_investment_rate_#.csv
In AMES, world prices of tradeable goods and services are exogenous, so they are not reported. Domestic prices are calculated endogenously by applying a markup. "Basic" prices of goods and services are calculated as a trade-weighted average of the world and domestic price. Domestic and basic prices are reported in the files:
domestic_prices_#.csv
basic_prices_#.csv