Experimental Data for GHD Computation under Updates

Georg Gottlob, Matthias Lanzinger, Davide Mario Longo, Cem OkulmusZenodo 2022 · DOI

Abstract

<strong>Full raw data and other data on the experiments</strong> The following is an overview on the contents of the data that<br> is provided here to allow third parties to analyze our experimental<br> data as well as give them all the tools needed to replicate them. <br> <strong>Naming of the update operations:</strong> We initially worked with the assumption of being in a database<br> context, with conjunctive queries as the example application. Thus<br> the names of updates was different when we ran the experiments.<br> Instead of speaking about adding or removing constraints, the term<br> 'relation' was used. Instead of trying to rename all our experiment<br> files, we simply provide a sort of translation table here: AddRel AddConstr DelRel DelConstr UnJoin* DelEq Join* AddEq UnConst AddVar * (in a database setting, we 'join' two relations) Note that in the paper we consistently use the names on the right,<br> while here in this repository the terms on the left will be used<br> throughout. <br> <strong>Contents</strong> rawdata_from_condor<br> <br> This contains all the output from our experiments on the 44950 instances<br> of our synthethic update testset. For each instance of this set, there are<br> four files in the subfolder 'tosolve': <br> 1) .sh<br> The script used to run it on our test machine.<br> 2) .error<br> Where the output from stderr was saved.<br> 3) .log<br> Where any log output from HTCondor itself was saved.<br> 4) .out<br> The actual stdout output from our test, which was accessed<br> to generate the parsed data.<br> original_decompositions_with_mutable_subtree<br> <br> This contains the 44950 GHDs that were used for our update tests.<br> These are _not_ GHDs of the updated hypergraphs, but instead are<br> GHDs for the original hypergraph from HyperBench. They are in a<br> JSON format, with indications with nodes form the minimal mutable<br> subtree. The name of the test instances as well as these<br> decompositions is constructed by taking the name of the original<br> hypergraph from HyperBench, and adding a suffix which indicates<br> the kind of update that was performed (the naming scheme for the<br> update operations was discussed earlier in this file).<br> update_instances<br> <br> The 44950 test instances that were synthetically generated, as<br> described in the paper. The naming scheme is the following: we start<br> with the name of the original hypergraph from HyperBench (these are<br> also included in repository) and adding a suffix indicating which of<br> the 5 elementary modifications we performed, followed by a number to<br> distinguish updates of the same kind.<br> HyperBench<br> <br> For convenience, we also provide here the hypergraphs from the<br> HyperBench dataset. These graphs and more info can be found under<br> its URL: http://hyperbench.dbai.tuwien.ac.at/<br> scripts<br> <br> This contains the scripts needed to generate the parsed data from<br> the raw data.<br> parsed_data.csv<br> <br> This contains the output from the aforementioned scripts and was<br> used to generate the statistics (in the form of a table and multiple<br> figures) we reported on in the paper.<br> stats.ipynb<br> <br> We provide as a JupyterLab notebook the queries we used to summarize<br> the run time for each method. The notebook allows accessing the<br> contents of the parsed data as a standard relational database and<br> SQL queries, thus allowing to easily produce more statistics on the<br> data, beyond what was provided in the paper. <strong>Links to source code of used decomposition methods: </strong> htdLEO: https://github.com/ASchidler/htdsmt BalancedGoUpate: https://github.com/cem-okulmus/BalancedGoUpdate
BibTeX
@inproceedings{gottlob2022experimental,
  author = {Georg Gottlob and Matthias Lanzinger and Davide Mario Longo and Cem Okulmus},
  title = {{Experimental Data for GHD Computation under Updates}},
  booktitle = {Zenodo},
  year = {2022},
  doi = {10.5281/zenodo.6481125},
}

Loading related content...

Figure