IBPSA Project 1 - WP3 meetings
Ina De Jaeger
Dear WP3 colleagues,
Thank you for filling in the Doodles. They have decided:
- The Buildings meeting will take place Monday, October 28, 5 pm Brussels time. I have sent out an invitation to those who were available in the Doodle. Everyone is more than welcome to join. The meeting details (how to connect? + agenda) can be found here.
- The Coordination meeting will take place Monday, November 18, 5 pm Brussels time. I have sent out an invitation to those who were available in the Doodle. Everyone is more than welcome to join. The meeting details (how to connect? + agenda) can be found here.
- The Network meeting: will be determined soon !
If you want me to forward the meeting invite, just drop me an e-mail.
I hope you are all well.
Have a nice weekend!
All the best
Ina
|
|
Ina De Jaeger Built Environment & Smart Cities, EnergyVille | Building Physics Section, KU Leuven | A: Kasteelpark Arenberg 40, box 2447, BE-3001 Leuven | P: +32 16 372 413 |
|
Hicham JOHRA
Hi everyone,
It took me a bit more time than expected, but I managed to implement everything we’ve discussed at the last WP.3 meeting for the Python tool to compare the data of a case with all the data from all the cases stored on Github.
I’ve tried to make the interface fairly straightforward with info messages.
Give it a try. Just run the python module enclosed and follow the steps. If you want to upload your data, do so (provided that it has the right format) or use the data file enclosed, or run the comparison procedure without your own data.
You can choose to add or remove different KPIs.
Only 4 are visible now (and only 3 actually computed) but it is fairly simple to implement new ones in the code.
It is also very flexible interface to integrate new cases with different parameters or more parameters, we just have to generate a parameter file .txt, and place it on the github folder with the right name, and add this name in the right list of the code.
The interface also allows to quickly check if a data file is valid according to the case that it is supposed to simulate.
It also gives the number of valid data files for a given case that are on the github folder (and I can see that some data files that have been uploaded there do not have the correct sampling time, which is 10 min interval, and are therefore considered invalid for the analysis).
The core of the interface and most of its calculation functions are done. What is missing is the online help file, the calculation of R2 (and eventually some other KPIs like Root Mean Squared Logarithmic Error, according to what people want). The graphs and pdf report generator output are not ready yet.
At the end of the procedure, it should generate a result table and few bar plots. If it does not, then check the “df_result” dataframe in the workspace that summarizes all the results of the DESTEST comparison procedure.
Please give it a try and let me know if it works or if it fails to deliver the “df_result”.
I will more actively write the paper for Building Simulation explaining this DESTEST comparison tool and call for a short meeting of W.P.3 network before Xmas.
Thanks