wateRinfo package #255
wateRinfo facilitates the access to a variety of environmental water-related data about Flanders (Belgium) available on waterinfo.be, a website managed by the Flemish Environmental Agency.
data retrieval: because the package provides R access to environmental data (water level and tidal information, discharge data, water quality variables, meteorological variables...) provided by the Flemish Environmental Agency. Without the package, people need to download data manually using the web interface or using the available API calls. The web interface, although evolving, does not support downloads for a list of stations and variables leading to lots of clicks to download data. The API documentation is limited and only available in Dutch. The package overcomes these limitations, supporting downloads of multiple variables and stations more efficiently.
Anyone interested in using the water-related data from waterinfo.be, for example hydrologists and ecologists. Scientific applications are ranging from environmental prediction and risk assessment modelling studies (e.g hydrological flood models) to ecological studies requiring environmental data.
To our knowledge, there is no other R package to retrieve data from waterinfo.be. Although some loose scripts have been circulating among researchers from different institutes, this package aims to support a collaborative/community effort.
Confirm each of the following by checking the box. This package:
Good point. I'll try to provide an overview of the information I'm aware of. Notice that we have for water-related issues different governmental levels: national level and regional levels, i.e. Flanders, Walloon and Brussels). Rivers are monitored on regional (and subregional) level:
Notice that we opt for GBIF to publish species occurrences related to water, e.g. fish, so this data is covered by the rgbif package.
As such, to my knowledge, there is no direct other source of data to incorporate. I'm adding @WillemMaetens (VMM), @pietercolpaert (open knowledge), @Sachagobeyn, @bartpannemans, @pjhaest to the discussion, maybe they know about other initiatives?
With respect to water quality data hosted by the VMM (Flanders): as I am currently aware, they are not available open source. They can be requested via the VMM website, however from my own experience, requests are often declined. An alternative for this type of information is the European Union Open Data Portal, in which ecological water quality, and a number of physico-chemical water quality variables are reported. For the latter, it should noted that data are only available for a few years. In addition, the datasets are not 'callable' as they are in waterinfo (zips hosted on the website).
Thanks for the information.
With respect to the issues with the example and unit tests: Apparently, the unique time series identifiers have changed on the waterinfo.be side for a subset of stations. I was expecting them to be stable, but they are not. I'm doing the best I can to fix these issues asap.
One of the elements missing in the unit testing is the XML-parsing of a specific type of error from the serverside at waterinfo.be (eventhough requesting
With respect to the Tex errors, I'm not completely aware on how to tackle this. I tried to execute
I get the warning:
However, when I do check the logs and output in the
And the output looks like: wateRinfo-manual.pdf Another file,
Any input on how to tackle this issue is certainly welcome.
Please check off boxes as applicable, and elaborate in comments below. Your review is not limited to these topics, as described in the reviewer guide
The package includes all the following forms of documentation:
I don't think this is the case???
Final approval (post-review)
Estimated hours spent reviewing: 3
Holy cow, nicely done! Followed the recommendations to the letter. A very minor suggestions...none of them are mandatory changes:
Maybe you could add a description to the
Seems like this package would be a prime candidate for language translation. The little bit I looked into this awhile back, I think this SO post could get you started:
I wonder if
There are a lot of Factors returned, but a character vector now-and-then. Factors work great for the plots, maybe not so great for other purposes. I'd double-check that the columns that are coming back as Factors make sense to be factors.
It might be useful to the user to offer a way to see or save the URL/GET query of the retrieved data. I'm not sure the best way, one could just be a message when retrieving, one way could be to add it to the attribute of the data.frame. (for the user, they may need to use it to cite the data, for you, you might need to use it when troubleshooting problems)
All-in-all, nice job!
@ldecicco-USGS thanks for your time and effort, it is really appreciated and provided very useful input. I have made changes and responded to your comments/suggestions below, quoting the commit the change was made in.
Response: As we are not controlling the server-side error messages neither the dutch fields embedded in the waterinfo.be website, we would have to do translations against a moving target that we do not control. Something to keep in mind if documentation of the API service improves, but we would keep this currently out of scope.
Response: Indeed, as such users will be able to iterate over these frequencies or subset them. Adapted (and adapted test), ropensci/wateRinfo@6023ca1
Response: Checked this and the main issue was the
Response: Available to the user, as it iw useful to have a check or control on the so-called
Response: Great suggestion! I found a possible solution to add the URL of the response as a comment attribute to the returned
As I used
I hope these adaptations incorporates your suggestions? Any suggestions are welcome.
Congrats @stijnvanhoey, your submission has been approved!
Welcome aboard! We'd also love a blog post about your package, either a short-form intro to it (https://ropensci.org/technotes/) or long-form post with more narrative about its development. ((https://ropensci.org/blog/). If you are, @stefaniebutland will be in touch about content and timing.
Thanks for the approval. Great news. I'll to the transfer as soon as possible.
@stefaniebutland, I've written a post/story about using the wateRinfo package to download tidal data to study eel behaviour. It is not a short-form intro or a narrative about the development, but would it also be an option as blog post? (notice, the current version is a draft, so text-wise not polished; but the storyline is there already)
@stijnvanhoey Your draft post will be great for the rOpenSci blog! Such a good weaving of practical and very engaging story with package use case. That video is incredible. We can publish on Jan 15, 2019.
Please submit your draft as a pull request by Jan 8 according to the instructions here: https://github.com/ropensci/roweb2#contributing-a-blog-post. That will give me an opportunity to review prior to publication.
Looking forward to this one!
Not a problem @stijnvanhoey. We're very happy to host your post when it is ready.
If you submit a pull request on Monday I'll do my best to do a final review asap after it's in. My responses might be delayed, however, because I'm involved in a workshop next week. The post is in very good shape so I don't expect it will need many changes after my review.
If you're not able to submit on Monday, then submit when ready and we can assign a different publication date.
rOpenSci will tweet about the post and you are welcome to suggest any tweet-content, preferred images or people who should be tagged and I will work to incorporate them.
Thank you for getting in touch!