Isabelle Weindl ([email protected])
- 1 Introduction
- 2 Model-internal R-scripts for output analysis
- 3 Automated model validation
- 4 Interactive scenario analysis
- 5 Analysis of outputs with the magpie4 library
- 6 Analysis of outputs with the gdx library
- 7 Lessons learned
After having successfully started and accomplished a simulation run, the next step is to evaluate the simulation results. In case you have not yet conducted an own MAgPIE simulation or your simulation is still running, you can download model runs produced with the master branch of the GIT repository (https://github.com/magpiemodel/magpie/tree/master) and copy the folders containing the simulation results into the output folder of your local version of the MAgPIE model. The runs can be downloaded from: http://bit.ly/MAGPIE4.
There are several ways to assess and evaluate MAgPIE results. This tutorial gives an overview on different tools and options that can be used to analyse model outputs.
For each simulation, results are written to a folder that is created automatically as a combination of the model title and other information (as defined in the default.cfg via the command cfg$results_folder) inside the output folder of the model.
The goal of this exercise is to use several tools for output analysis. After completion of this exercise, you’ll be able to:
- Use model-internal R-scripts for output analysis.
- Know where to find the automated validation pdf and how it is structured.
- Use the evaluation tool appResultsLocal of the library shinyresults.
- Use the magpie4 library for output analysis.
- Analyse outputs with the gdx library.
In the file “config/default.cfg”, it is possible to indicate which R-scripts are executed for output analysis after a model run is finished. Scripts evaluating single runs are stored in the folder scripts/output/single, while the folder scripts/output/comparison contains scripts that compare model output across several runs. In the default MAgPIE configuration, the scripts rds_report (to be used in appResultsLocal; see explanations below), validation and interpolation are selected via cfg$output:
cfg$output <- c("rds_report","validation","interpolation")Output scripts that are included in the folders scripts/output/single and scripts/output/comparison can also be executed via command window. To do so, windows users can open a command line prompt in the MAgPIE model folder by using shift + right click and then selecting open command window here option.
In the command prompt, use the following command:
Rscript output.RYou are now asked to choose the output mode: 1: Output for single run 2: Comparison across runs
In both cases, you can choose from the list of available model simulations, for which runs you want to conduct the model output analysis:
In the next step, you can interactively indicate which model-internal output scripts you want to execute:
The last step is to select the run submission type, e.g.“Direct execution”:
Now, the selected script is executed. After completion, the results are written in the respective folder of the simulation run inside the output folder of the model.
Execute the model-internal output script report.R via command
window. This script collects the results of several report-functions -
that calculate many key output variables like Production, Land Use or
Yields - and writes them into one mif-file that can be read with text
editors. Please refer to the vignette("mif") of the package mip
(model intercomparison plots) to learn more about the mif format.
The automated model validation is an important example of output analysis based on model-internal scripts (see section 2). If the validation script is executed (either by selection via cfg$output as explained in 2.1. or by execution via command window as explained in 2.2.), a standard evaluation pdf is created that validates numerous model outputs with a validation database containing historical data and projections for most outputs returned by the model, either visually or via statistical tests. A standard evaluation PDF consists of hundreds of evaluation outputs and usually has a length of around 1800 pages. By evaluating the model outputs on such a broad level rather than focusing only on key outputs, it allows getting a more complete picture of the corresponding simulation. As an example of such validation files, you can download the evaluation documents produced for all runs shown in the MAgPIE 4 framework paper (https://doi.org/10.5281/zenodo.1485303).
The table of contents of the validation pdf gives a good overview about the breadth of model outputs that can be simulated with a MAgPIE standard simulation, even though the validation pdf only shows a subset of possible model outputs:
Open a validation pdf (either in a folder containing your own simulation results or the downloaded MAgPIE simulation runs, or the validation files that show results of the simulation runs used for the MAgPIE4 paper) and
- make yourself familiar with the structure of the document and the hierarchy of outputs as displayed by the table of contents as well as
- have a look at some figures displaying model outputs of your interest.
The automated model validation is a good tool for visually evaluating a broad range of model outputs. However, comparison between model runs, i.e. between different scenarios, is rather difficult and inconvenient if the model results are scattered across different large PDF files.
To overcome this issue, we developed the interactive scenario analysis and evaluation tools appResultsLocal (and appResults for the use within the PIK network) as part of the library shinyresults (https://github.com/pik-piam/shinyresults), which show evaluation plots for multiple scenarios including historical data and other projections based on an interactive selection of regions and variables. You can use this tool by running the following R command in the main folder of your model, which will automatically collect all runs in the output folder and visualize them:
shinyresults::appResultsLocal()This command opens an interactive window, where you can select the simulations that you want to evaluate.
You can use filters to select a subset of all runs stored in the output folder of the model, for example by searching for runs that have been finished at a certain day or by searching for keywords in the title of the simulation runs:
Choose title as filter and select 2 simulations that are stored in the output folder of your model, e.g. the SSP2 and SSP5 simulations from the downloaded set of MAgPIE runs:
After having selected the subset of runs that you want to analyse, click the button Load selection. Now, you can click on the tab LinePLot. You will then see on the right hand side line plots showing the development of population for historical and future time steps for all model regions and on the global scale:
Now, choose a variable of your interest, either by scrolling through the drop-down menu or write a key word in the input field, e.g. “cropland”, to reduce the options in the menu.
Now, make yourself familiar with the features of the app. You can for example select a subset of regions or a subset of time steps for which the results should be plotted. Moreover, you can free the y-axis, include or exclude validation data (if available) and download the plot.
If you want to go beyond visual output analysis and predefined output evaluation facilitated by scripts in the model folders scripts/output/single and scripts/output/comparison, you can use the functionality of the R package magpie4 (https://github.com/pik-piam/magpie4). This library contains a list of common functions for extracting outputs from the MAgPIE model which are also the basis for the generation of the automated validation pdf. For a quick overview on the functions which are included in the library, you can scan the folder magpie4/R. The following figure shows a subset of R-files included in magpie4/R:
For making yourself familiar with this library, you can open a R/RStudio session and set the MAgPIE model folder as working directory. This can be done by using the following command:
setwd("/path/to/your/magpie/model/folder")Then, load the library and call the help pages:
library(magpie4)
?magpie4You can click on the index and search for interesting functions, e.g. carbonstock, and read the respective help page:
Apply the function carbonstock for calculations in R, by using the file “fulldata.gdx” in the folder of a model simulation run as input for the function.
- Use the default settings of the arguments of the function
- Change the arguments of the function, e.g. set the level from “cell” to “reg”.
The gdx library (https://github.com/pik-piam/gdx) allows for directly accessing objects contained in the fulldata.gdx via the function readGDX. A pragmatic way to learn how to use this function for the extraction of interesting information from the fulldata.gdx is to open R-files of the magpie4 library within Rstudio. Most of the magpie4 functions make use of readGDX.
In the function carbonstock of the magpie4 library, we see several instances where readGDX is used, e.g.:
a <- readGDX(gdx,"ov_carbon_stock",select=list(type="level"),react="silent")
sm_cc_carbon <- readGDX(gdx,"sm_cc_carbon2",react = "silent")
ov_land <- readGDX(gdx,"ov_land",select = list(type="level"))
fm_carbon_density <- readGDX(gdx,"fm_carbon_density")[,t,]It is possible to extract various GAMS objects like “sets”, “equations”, “parameters”, “variables” and “aliases” with readGDX.
With the argument select=list(type=“level”), you can select the levels of endogenous variables, with select=list(type=“marginal”) you can extract the marginal values of these variables.
- You have an overview on several tools analysing MAgPIE outputs.
- You know how to start model-internal R-scripts.
- You know how to generate and where to find the automated validation pdf and how it is structured.
- You know how to start the evaluation tool appResultsLocal of the library shinyresults and how you can visualized MAgPIE results with this tool.
- You know the general functionality of the magpie4 library and how to use some of the functions.
- You know the general functionality of the gdx library.