Q&A Webinar

Below you will find the questions and answers that emerged during our webinar on July, 11th.
Note:

This page will be taken down in the near future!
Long-term relevant questions will be included in the FAQ section of the GAMS MIRO documentation.

General

I downloaded the latest GAMS distribution from www.gams.com/download but I can not see any MIRO buttons in GAMS Studio.

During the beta phase a special GAMS distribution will be available which contains GAMS MIRO. The regular GAMS distributions are completely independent of this. The MIRO buttons are only available in the version of GAMS Studio that ships with the MIRO BETA version.

Are the MIRO buttons also included in the GAMSIDE?

No, the GAMSIDE is no longer maintained. If you want to start MIRO from the GAMSIDE or the command line, you have to set the command line option miro=launch. For the configuration mode you additionally have to add the option miromode=2. You can find more information about launching MIRO and starting the configuration in the MIRO docs.

Features

Can I use notebooks in the Hypercube mode in MIRO 1.0?

We plan to allow notebook technology also for the Hypercube mode.

Will the MIRO server version only be available as-a-service or can I use it on my own server?

As things currently stand, both variants will be possible.

With MIRO, can I continue using the GAMS Python API to drive the GDX generation and GAMS modeling?

Here, we have to distinguish between different things:
  • Embedded Python code can be used in your GAMS models without any restrictions.
  • The GAMS Python API gives Python the control over running GAMS. MIRO is a different way to run GAMS models so these two things are by nature incompatible.
  • The MIRO back end (not the web front end) that executes your GAMS models will have a REST API and can therefore be called from a Python script and will complement and extend the already available GAMS Python API.

Does MIRO only support those graphics which are available in the configuration mode?

No. You can write your own Custom Renderers in R using MIRO's R API.

In GAMS MIRO, after setting up a new configuration of a Table (i.e., setting up Rows and columns), can I export that table to Excel or Word in the same format?

Tables can be exported as Excel files in the format in which they are displayed in MIRO. Just click on Scenario in the upper right corner and then export. There you can choose between GDX and XLS format. This exports not only one table but the whole scenario data. Read more about saving scenario data in the MIRO docs.
If you want to export individual tables, this might be a reasonable feature request. Currently it is only possible in this way:
  • For the output tables (which can also be configured as input tables), buttons can be added in the table settings of the configuration mode. These allow the export of individual tables (Attention: Only the currently visible table page will be exported).
  • With the regular input tables you can mark all entries in a table, copy them to the clipboard and paste them into an Excel sheet.

How is the performance with large GDX files and parameters?

Performance and large are both very imprecise terms. We have tested MIRO with datasets up to 1 million records. Data import at this instance takes about 5 seconds and the data export about 7 seconds on a computer with an Intel Quad-Core i7 with 2.8 GHz.
Please note that the default tool for rendering input tables loads the entire dataset into the user's browser. This can be problematic if you have clients working on low performance computers and/or old browsers. Thus, we have added a "big data" version for input tables that supports pagination and therefore does not depend as much on the hardware of the client. These "big data" tables are not yet included in the current beta version, but will be available in upcoming releases.

In your examples you changed parameter values - can you also change set global variables?

In addition to GAMS symbols like parameters and scalars, command line parameters (GAMS Options and Double dash parameters) can also be set from within MIRO. If a GAMS option or a double dash parameter is set via MIRO, your GAMS model will automatically be called with these command line parameters. Those values the user selected are therefore available to the GAMS model at compile time.
Example:
If we set the double dash parameter --numberPeriods to 11 and for the GAMS option mip we select CPLEX from within MIRO, then the GAMS model is run with:

gams <modelname>.gms --numberPeriods=11 MIP=CPLEX
Read more about the usage of command line parameters in the MIRO documentation

Is it possible to hide the solving process for the user, e.g. display only the result message?

Yes. You can configure the interface to show only a status message and hide the GAMS log and listing (lst) file. In the configuration mode you can find this option in the general settings.

Do you have a plan for equipping GAMS MIRO with econometric tools (forecasts) for stock analysis?

No. We do, however, plan to integrate "custom output sheets" that allow you to execute your own code. These can be used to perform statistical analyses among others.
Perhaps this can be part of your GAMS model that either includes this logic as (Python) embedded code or calls the required tools via file exchange and processes. Usually for security reasons the GAMS program is limited in its action to call programs and execute embedded code, but these security settings can be relaxed by the app maintainer.

If I got it right, each view is related to one parameter, or can you have more views per parameter?

For each Parameter you always have the tabular view and a graphical representation available (side by side or on single sheets). If you want to have several graphics for one parameter, this is possible via the R API.
For the 1.0 release we plan to allow the usage of advanced analysis capabilities, e.g. Jupyter notebooks, to have multiple graphics for a single symbol or graphics that combine data of multiple symbols. This will require some Python or R programming

Do you plan for spanish language support?
Which languages do you plan to support?

GAMS MIRO has the possibility to change the preferred language. This results in MIRO including the corresponding language package. Language packages are JSON files in which all language-specific components are defined. To create a new language package, this JSON file only needs to be translated. We are happy when users want to create new language packs. If you would like to contribute your language file, so that other people can profit from your work as well, send us an email: miro@gams.com.

Development

What is the rough schedule of the server infrastructure (GAMS REST API) in your roadmap?

We are working hard on the server version and plan a release later this year.

Is it possible to use a web server in order to publish results even if it is not official yet?

Yes, if you would like us to provide you with an account on one of our servers for testing MIRO, please contact us. Please note that we operate a GAMS MIRO gallery in which we present some exemplary models. If you want to add your model there, please reach out!