Revision DesInventar metadata description and proposal loss group

[bennyistanto]

This PR contains:

• Revision of DesInventar markdown's dataset long description in single paragraphs, which replace into shorter description and grouping based on hazard_type
• Proposal for Improving (individual) metadata page #115

[github-actions]

📊 PR Sync Results

📝 Dataset Changes

• No dataset changes detected

🧮 Vectors

• ⏭️ No vector updates needed

✅ Validation

• ⏭️ Validation skipped

🧪 Search Tests

• ❌ 0/1 tests passed

❌ Failed: "vulnerability flood"

Description: Test adapted from comments on Issue #31

Expected includes:

Dataset	Expected	Found	Rank	Score
rdls_vln-jrc_glb_flood	✅	✅	3	0.4786
rdls_lss-gmb_floodcoastres	✅	✅	15	0.4246
rdls_vln-undrr_glb_floodinfrastructure	✅	✅	23	0.4113
rdls_vln-wb_sfrarr_flood	✅	✅	21	0.4148
rdls_hevl-kentmrwcities_nairobi	✅	✅	26	0.3985
rdls_hevl-kentmrwcities_nakuru	✅	✅	14	0.4268
rdls_lss-tmrwcities_daressalaam	✅	✅	28	0.3821
rdls_hevl-bgdtmrwcities_chattogram	✅	✅	34	0.3761
rdls_hevl-ecutmrwcities_quito	✅	❌	64	0.2947
rdls_hevl-npltmrwcities_rapti	✅	❌	66	0.2813
rdls_hevl-psetmrwcities_nablus	✅	❌	75	0.2578
rdls_lss-tmrwcities_coxsbazar	✅	✅	37	0.3701
rdls_lss-tmrwcities_daressalaam	✅	✅	28	0.3821

Missing includes: rdls_hevl-ecutmrwcities_quito, rdls_hevl-npltmrwcities_rapti, rdls_hevl-psetmrwcities_nablus

Top 5 actual results:

Rank	Dataset	Score
1	rdls_hzd-wb_sfrarr_floodscenario	0.5040
2	rdls_hzd-fathom_glb_floodoac	0.4961
3	rdls_vln-jrc_glb_flood	0.4786
4	rdls_vln-wb_sfrarr_earthquake	0.4706
5	rdls_hzd-gfdrr_afg_flood	0.4625

[bennyistanto]

@matamadio Thanks for the reminder during the recent call. I apologize, I don't know if the rdl-02.yml and rdl-03.yml are the schema in JKAN.

As we will not touch the schema, I revised the approach for visualising the Loss group, and only dealing with html and css (edit and add new file). So far looks great. See below and looking forward to receive your feedback.

Compare to Loss dataset with hazard_type less than 3

[matamadio]

I like the grouping approach 👍

[lydiascarf]

@bennyistanto this looks good to me, except for two things:

• when i rebase with git rebase -i origin/rdl-0.3, i realize this PR is reviving the main.css file we deleted for the redesign. if i then, delete the file it mostly looks good, but the collapsible loss group sections are no longer correctly styled. could you please take a look at what happened there? i can help, but i have a lot to do with the urban data tracker ASAP, so i'll wait for you to let me know if you need help
• after rebasing, can you please run the Python script again with python3 main.py --markdown because i'm seeing changes after rebasing that won't be triggered by CI

[matamadio]

when i rebase with git rebase -i origin/rdl-0.3, i realize this PR is reviving the main.css file we deleted for the redesign. if i then, delete the file it mostly looks good, but the collapsible loss group sections are no longer correctly styled. could you please take a look at what happened there? i can help, but i have a lot to do with the urban data tracker ASAP, so i'll wait for you to let me know if you need help

We merged main.css into https://github.com/GFDRR/rdl-jkan/blob/rdl-0.3/css/modern-theme.css, it should be sufficient to paste new styling there.

[bennyistanto]

@bennyistanto this looks good to me, except for two things:

• when i rebase with git rebase -i origin/rdl-0.3, i realize this PR is reviving the main.css file we deleted for the redesign. if i then, delete the file it mostly looks good, but the collapsible loss group sections are no longer correctly styled. could you please take a look at what happened there? i can help, but i have a lot to do with the urban data tracker ASAP, so i'll wait for you to let me know if you need help
• after rebasing, can you please run the Python script again with python3 main.py --markdown because i'm seeing changes after rebasing that won't be triggered by CI

@lydiascarf I have deleted the main.css, and run python3 main.py --markdown as you suggested.

[lydiascarf]

just rebased to address conflicts that have cropped up in the last couple months!

[lydiascarf]

the other datasets have licenses in the code form. the full license names live in _data/rdl-license-name.yml

[bennyistanto]

From the filename, seems this dataset came from first iteration when I worked on hdx, after that so many change in the schema and the HDX pipeline too.

[lydiascarf]

how should loss_count and loss_groups render? i want to add this to the schema and the UI templates

[bennyistanto]

@matamadio what is the best way to handle many errors due to very old version metadata. I dont think its a good idea for @lydiascarf to fix this various problem in v0.3 json datasets.

How about I share some of the example JSON v1.0 from various project/case, so Lydia can directly working using correct metadata for rdl-1.0 ?

[matamadio]

@matamadio what is the best way to handle many errors due to very old version metadata. I dont think its a good idea for @lydiascarf to fix this various problem in v0.3 json datasets.

How about I share some of the example JSON v1.0 from various project/case, so Lydia can directly working using correct metadata for rdl-1.0 ?

Agree, let's focus on 1.0 release which will replace most/all previous datasets.
On that basis, let's provide instructions on how to group/visualise the metadata for multiple hazard/losses.
I'll open a new issue for that.