Single Pkg, Examples, small var cost for storage #5
Conversation
YoungFaithful
commented
Apr 23, 2019
YoungFaithful
commented
- StatsKit is not included, instead CSV, DataFrames, and StatsBase (for faster startup)
- Storage technologies now have a very small variable cost (otherwise charging and discharging can happen in times of too high renewables)
- Some correction in documentation
- Examples updated
| @@ -4,7 +4,7 @@ bat_e,all,2016,fix,|,3020,0 | |||
| bat_e,all,2016,var,|,0,0 | |||
| bat_in,all,2016,cap,|,280000,0 | |||
| bat_in,all,2016,fix,|,2800,0 | |||
| bat_in,all,2016,var,|,0,0 | |||
| bat_in,all,2016,var,|,0.01,0 | |||
There was a problem hiding this comment.
small variable cost to prevent charge and discharge at the same time in times of renewable overproduction (Excess energy was able to be charged and discharged in the same hour without additional costs)
There was a problem hiding this comment.
Nice, I always was wondering if degeneracy was occuring.
There was a problem hiding this comment.
Just as a note: Did you add a sentence about this to the documentation?
There was a problem hiding this comment.
Good thought. Added it ;)
There was a problem hiding this comment.
Some additional comments.
Ready to be reviewed @holgerteichgraeber
| @@ -44,7 +44,7 @@ Pkg.add(PackageSpec(url="https://github.com/holgerteichgraeber/ClustForOpt.jl", | |||
| ``` | |||
There was a problem hiding this comment.
Above, there is a inconsistency .jl. JuMP.jl vs JLD2 for example.
| @@ -4,7 +4,7 @@ bat_e,all,2016,fix,|,3020,0 | |||
| bat_e,all,2016,var,|,0,0 | |||
| bat_in,all,2016,cap,|,280000,0 | |||
| bat_in,all,2016,fix,|,2800,0 | |||
| bat_in,all,2016,var,|,0,0 | |||
| bat_in,all,2016,var,|,0.01,0 | |||
There was a problem hiding this comment.
Nice, I always was wondering if degeneracy was occuring.
| @@ -4,7 +4,7 @@ bat_e,all,2016,fix,|,3020,0 | |||
| bat_e,all,2016,var,|,0,0 | |||
| bat_in,all,2016,cap,|,280000,0 | |||
| bat_in,all,2016,fix,|,2800,0 | |||
| bat_in,all,2016,var,|,0,0 | |||
| bat_in,all,2016,var,|,0.01,0 | |||
There was a problem hiding this comment.
Just as a note: Did you add a sentence about this to the documentation?
|
|
||
| # Desing with clusered data and operation with ts_full_data | ||
| # First solve the clustered case | ||
| design_result = run_opt(ts_clust_data.best_results,cep_data,optimizer;descriptor="design&operation", co2_limit=50) | ||
| design_result = run_opt(ts_clust_data.best_results,cep_data,optimizer;co2_limit=50) |
There was a problem hiding this comment.
Just to check if I understand this: Basically how I see it is that the descriptor moved into the variable description? Great, that makes a lot of sense to me.
There was a problem hiding this comment.
yes exactly. First users were getting a little confused, what the descriptor does...
| # load cep-data | ||
| cep_data = load_cep_data_provided(state) | ||
|
|
||
| ## CLUSTERING ## | ||
| # run aggregation with kmeans | ||
| ts_clust_data = run_clust(ts_input_data;method="kmeans",representation="centroid",n_init=5,n_clust=5) # default k-means make sure that n_init is high enough otherwise the results could be crap and drive you crazy | ||
| ts_clust_data = run_clust(ts_input_data;method="kmeans",representation="centroid",n_init=10000,n_clust=5) # default k-means make sure that n_init is high enough otherwise the results could be crap and drive you crazy |
There was a problem hiding this comment.
How about keeping it at 100 or 1000 and mentioning in a comment that 10000 should be used for real applications?
|
Implemented requests. |
