Update coupling tutorial: how to specify MAgPIE scenarios using the p…

…ath to a config file in the coupled config.

tutorials/04_RunningREMINDandMAgPIE.md

@@ -100,13 +100,17 @@ See comments in the head section of the file. Most importantly you need to provi
 Required as `path_settings_coupled` is a file from [`./config/`](../config) that starts with `scenario_config_coupled*.csv` and provides some extra information for coupled runs (e.g. which run should be started, specific MAgPIE configurations), with one scenario per row and settings on the columns. `path_settings_remind` is a normal `scenario_config` that defines all other REMIND settings, as explained in [`03_RunningBundleOfRuns`](./03_RunningBundleOfRuns.md). Every scenario to be run must be present in both files.
 
 All the columns must be present in the `scenario_config_coupled.csv` file, but most of them can be left blank. The required ones are:
+
    - `title`: The name of the scenario, must be unique and match the `title` column in REMIND's `scenario_config.csv`
    - `start`: Defines if a scenario run should be started (1) or not (0). Overrides whatever is set in REMIND's `scenario_config*.csv`. If you have an unfinished coupled run, it will automatically try to continue from the last coupling iteration (i.e. REMIND or MAgPIE run), if it is not finished yet (i.e. reached `max_iterations`).
    - `qos`: The SLURM qos the coupled runs should be submitted to, such as `priority`, `short`, `medium`, `long` or `standby` on the PIK cluster. If you use `auto`, the qos is set to `priority` if an empty slot exists, and `short` otherwise. This is also the default if no `qos` is set. Currently, there is no default support for running a coupled run locally.
-   - `magpie_scen`: A pipe (`|`) separated list of configurations to pass to MAgPIE. Each entry should correspond to a column in [MAgPIE's scenario_config](https://github.com/magpiemodel/magpie/blob/master/config/scenario_config.csv), each one of them setting the multiple configuration flags listed on that file. The configurations are applied in the order that they appear. For example, to configure MAgPIE with SSP2 settings and climate change impacts according to RCP45 set `magpie_scen` to `SSP2|cc|rcp4p5`.
+
+   - `magpie_scen`: A pipe (`|`) separated list of configurations to pass to MAgPIE. Each entry should correspond to a column in [MAgPIE's scenario_config](https://github.com/magpiemodel/magpie/blob/master/config/scenario_config.csv), each one of them setting the multiple configuration flags listed in that file. The configurations are applied in the order that they appear. For example, to configure MAgPIE with SSP2 settings and climate change impacts according to RCP45 set `magpie_scen` to `SSP2|cc|rcp4p5`. To select scenarios from a different `scenario_config*.csv` file replace `magpie_scen` by the path to that file relative to MAgPIE's main folder, for example: `config/projects/scenario_config_PROJECT.csv`. The filename *must contain* the string `scenario_config`. You can also specify more than one column directing to a `scenario_config.csv` file. They will be evaluated in the order in which they appear in the `scenario_config_coupled.csv`.
+
    - `no_ghgprices_land_until`: Controls at which timestep in the MAgPIE runs GHG prices from REMIND will start to be applied. This essentially enables you to set whether or not (or when) GHG prices on land should be applied in MAgPIE. If you want MAgPIE to always apply the same GHG prices from REMIND, you should set this to a timestep corresponding to the start of your REMIND run, such as `y2020` to start in the 2020 timestep. If you want to disable GHG prices in MAgPIE, regardless of what REMIND finds, set this to the last timestep of the run (usually `y2150`). Values in between allow the simulation of policies where GHG prices are only applied in the land use sector after a certain year.
 
 Other, optional columns allow you to make a run start only after another has finished, set starting conditions, and give you finer control over which data is fed to MAgPIE.
+
    - `path_gdx`, `path_gdx_ref`, `path_gdx_refpolicycost`, `path_gdx_carbonprice`, `path_gdx_bau`, : Override these same settings in REMIND's `scenario_config`, see [`03_RunningBundleOfRuns`](./03_RunningBundleOfRuns.md) for a detailed explanation.
       - You can set these switches either to the full path of a `fulldata.gdx` file or simply to the name of another scenario in the file (without the "C_"!). So if you want a certain scenario (say `NDC`) to use as starting point the results of a `Base` scenario, you can simply set `path_gdx` to `Base` and it will automatically locate the appropriate `fulldata.gdx` in `Base`, for example `path_remind/C_Base-rem-x/fulldata.gdx`.
       - If you set any of these `path_gdx…` columns (or `path_mif_ghgprice_land`, below) with a scenario name such as `Base`, the coupling script will not start any runs that depend on an unfinished run, and automatically start them when that run finishes. So, in the example above, you can set `start` to `1` in both `Base` and `NDC`, `NDC` will only start *after* `Base` is finished. `NDC-rem-2` will start after `NDC-rem-1` and `Base-rem-2` are finished.