I've just started playing with CubeVizjs, and I would like to contribute some questions / comments on issues I stumbled upon. In general, they may be considered related to configuration parameters, that, as far as I could see, are not supported. My apologies in advance if this is not the case.
My use case
I'm using CubeVizjs on top of a SPARQL endpoint, giving access to ~60 qb:DataSet's, for a total of ~44,000 qb:Observation's.
Issues
IC validation
I had issues with IC-12 ("No duplicate observations") - the SPARQL endpoint takes ages to respond.
Also, I wonder whether it could be possible to enable / disable IC validation via the data_configuration parameters. As far as I can understand, this is currently not possible, and validation is carried out whenever the page is loaded and a qb:DataSet selected. This may be not always needed, e.g., if the data come from a SPARQL endpoint - in such a case, validation could be done only once.
"No measure selected" error in the log box
This occurrs to me when I switch to another qb:DataSet in the dropdown. I do select the relevant measure, but it seems that CubeVizjs does not detect it. NB: All datasets use just 1 measure (although it is not the same in all datasets).
I found a quick fix by making all select boxes "multiple" - now it works.
UI inconsistencies
It is unclear why the select boxes of attributes and dimensions have an "OK" button, which is missing from those concerning datasets and measures.
It may be worth having always such a button - which could be also a way to fix the aforementioned issue.
Default language of textual labels
My understanding is that CubeVizjs shows the first textual label found (rdfs:label, skos:prefLabel, etc.), irrespective of the language tag.
In my use case, I'm using WurVoc's OM vocabulary (for units of measurement), which includes textual labels in multiple languages. I'd like to get the English version of them, but I get instead the Dutch one - which seems to be the "first" returned by the relevant SPARQL query.
It would be useful to have a configuration parameter (maybe in ui_configuration) where I can state what is my preferred default language.
I've just started playing with CubeVizjs, and I would like to contribute some questions / comments on issues I stumbled upon. In general, they may be considered related to configuration parameters, that, as far as I could see, are not supported. My apologies in advance if this is not the case.
My use case
I'm using CubeVizjs on top of a SPARQL endpoint, giving access to ~60
qb:DataSet's, for a total of ~44,000qb:Observation's.Issues
IC validation
I had issues with IC-12 ("No duplicate observations") - the SPARQL endpoint takes ages to respond.
I wonder whether it would be possible to use an alternative, more efficient implementation of the relevant SPARQL query. I think an example is reported here: https://github.com/yyz1989/NoSPA-RDF-Data-Cube-Validator#performance
Also, I wonder whether it could be possible to enable / disable IC validation via the
data_configurationparameters. As far as I can understand, this is currently not possible, and validation is carried out whenever the page is loaded and aqb:DataSetselected. This may be not always needed, e.g., if the data come from a SPARQL endpoint - in such a case, validation could be done only once."No measure selected" error in the log box
This occurrs to me when I switch to another
qb:DataSetin the dropdown. I do select the relevant measure, but it seems that CubeVizjs does not detect it. NB: All datasets use just 1 measure (although it is not the same in all datasets).I found a quick fix by making all select boxes "multiple" - now it works.
UI inconsistencies
It is unclear why the select boxes of attributes and dimensions have an "OK" button, which is missing from those concerning datasets and measures.
It may be worth having always such a button - which could be also a way to fix the aforementioned issue.
Default language of textual labels
My understanding is that CubeVizjs shows the first textual label found (
rdfs:label,skos:prefLabel, etc.), irrespective of the language tag.In my use case, I'm using WurVoc's OM vocabulary (for units of measurement), which includes textual labels in multiple languages. I'd like to get the English version of them, but I get instead the Dutch one - which seems to be the "first" returned by the relevant SPARQL query.
It would be useful to have a configuration parameter (maybe in
ui_configuration) where I can state what is my preferred default language.