search

riatelab / magrit

Thematic cartography entirely in the browser ♥
https://magrit.cnrs.fr
GNU General Public License v3.0
128 stars 18 forks source link

Bad message wording when rows missing in dataset #123

Closed asitemade4u closed 8 months ago

asitemade4u commented 9 months ago

When joining a basemap and a dataset I get the message:

2024-02-14_15-17

Which is not relevant. One should possibly say:

The values on which to operate have to be unique.

Besides, it would be helpful to have a log of the duplicate rows.

asitemade4u commented 9 months ago

So, I checked the GISJOIN column in both the GeoJSON file and the CSV file. There are NO duplicates. What is true however, is that some rows are missing in the CSV file -- which makes sense as I have removed the NAs prior to exporting the CSV.

asitemade4u commented 9 months ago

Eventually, there were two no_data lines in the GeoJSON and they were the culprit.

mthh commented 9 months ago

Thanks for the report. Indeed, the wording of the message should be changed to add the missing "to" and remove the unnecessary "s". However I checked your data (from #122) and there are non-unique values in the GISJOIN field of the GeoJSON layer.

duplicate-rows

If you change one of these values (for example changing the first "G3600050nodata" to "G3600050nodata1") when preparing your data, you should be able to make this join.

a b c

But Magrit should probably not forbid joining data when rows are missing in the dataset, just consider them as NAs and depict them as such.

In your case, this is not a problem because you removed the missing rows yourself.

If you had left them in (with empty values in the join field), that would have been a problem (that's what's being discussed here, I think: https://github.com/riatelab/magrit/issues/30).

This is a limitation of our current join system. We're probably not going to improve it (just fix bugs if necessary) because we're developing a completely new version of the application, the join system having been refactored to be more flexible in these respects.

asitemade4u commented 9 months ago

Thank you for your prompt response. I am having a blast with Magrit, which is a great program. I discovered it by previously using mapsf, which is great too, and serves another kind of purpose. BTW, being French and living in the U.S. I am happy to help you with any translation...

mthh commented 9 months ago

I am having a blast with Magrit, which is a great program.

Nice to read this, thanks :)

BTW, being French and living in the U.S. I am happy to help you with any translation...

I'm taking note of your suggestion! If you're interested we'll let you know when we'll be making a beta of the new version available (in a few months' time), if you're available then, I'm sure you'll have some good suggestions for us in that area!

asitemade4u commented 9 months ago

I will be available for you, do not hesitate. I am testing Magrit in and out and am collecting all the (little) inconsistencies I find. Will send later.