Indicateurs (Indicators)

Indicators are not a separate data.gouv.fr object type: they are regular datasets that opt in via a tag and belong to a specific organization.

Selection

Configured in configs/ecospheres/config.yaml under pages.indicators of https://github.com/opendatateam/udata-front-kit/:

indicators:
  object_type: datasets
  universe_query:
    tag: ecospheres-indicateurs
    organization: 67884b4da4fca9c97bbef479
  filter_prefix: ecospheres-indicateurs

object_type: datasets — the indicators list is a filtered dataset search
universe_query: only datasets tagged ecospheres-indicateurs and belonging to organization 67884b4da4fca9c97bbef479 are returned.
filter_prefix: ecospheres-indicateurs is reused to:
detect a single dataset as an indicator client-side (dataset.tags.includes(filter_prefix)),
namespace the faceted filter tags (e.g. secteur), built as ${filter_prefix}-${filterId}-${valueId} (see details below).
Additional facets (secteur, date de mise à jour, etc.) are plain data.gouv.fr tags following that ecospheres-indicateurs-<filter>-<value> convention, configured under pages.indicators.filters in the same YAML.

Tag format for filters: `secteur` example

See https://github.com/opendatateam/udata-front-kit/blob/main/configs/ecospheres/config.yaml > pages.indicators.filters

- id: secteur
  ...
  use_filter_prefix: true
  values:
    - id: agriculture-forets-sols
      name: Agriculture, Forêts et Sols
    - id: alimentation
      name: Alimentation
    - id: batiment
      name: Bâtiment
    - id: dechets
      name: Déchets
  ...

For a dataset to appear under a given "Secteur" facet value, the dataset producer must add the corresponding tag ecospheres-indicateurs-secteur-<value.id> to the dataset, i.e. ${filter_prefix}-${filter.id}-${value.id}.

For "Agriculture, Forêts et Sols" that tag is ecospheres-indicateurs-secteur-agriculture-forets-sols.

A dataset can carry several ecospheres-indicateurs-secteur-* tags if it spans multiple sectors, in addition to the base ecospheres-indicateurs tag required for it to be picked up as an indicator at all (see universe_query above).

Native data.gouv.fr metadata used

See https://guides.data.gouv.fr/api-de-data.gouv.fr/reference/datasets

Standard dataset (DatasetV2) fields, displayed as-is:

Field	Used for
`title`, `slug`, `id`	Card/detail titles, routing, identifier display
`description`	Detail page body, card excerpt
`tags`	Indicator detection + faceted filters (see above), keyword list
`spatial.zones`	"Couverture géographique"
`temporal_coverage`	"Couverture temporelle"
`created_at`, `last_update`	"Date de création" / "Date de mise à jour"
`frequency`	"Fréquence de mise à jour"
`license`	Licence badge
`organization`	Sidebar attribution
`contact_points`	Array of attributions / contacts
resources (`main` group)	"Fichiers" tab, and viz data source (see below)
`extras` (generic)	Rendered in a raw "Extras" accordion for debugging

Custom metadata (extras)

Dataset extras — `ecospheres-indicateurs`

Schema: extras_schema.json

Key	Type	Required	Used for
`unite`	string	✅	Unit label (info panel, one-year value)
`mailles_geographiques`	string[]	✅	List of geo levels the indicator covers ("Mailles")
`axes`	object (`{ [axis]: string[] }`)	✅	Declares breakdown axes at dataset level
`calcul.responsable` / `calcul.methode`	string	✅	"Informations calcul" section
`sources[]` (`nom`, `url`, `description`, `producteur`, `distributeur`, `plage_temporelle.start/end`)	array	✅	"Sources" tab
`enable_visualization`	boolean	✅	Shows/hides the "Prévisualisation" tab and chart
`summable`	boolean	optional	Whether axis values can be summed/grouped in the chart
`y_start_at_zero`	boolean	optional	Chart y-axis behavior
`ignore_format_big_number`	boolean	optional	Skip k/M/Md number formatting
`next_expected_update_quarter`	string	optional	"Prochaine mise à jour attendue"

Resource extras — `ecospheres-indicateurs`

Schema: resource_extras_schema.json

Applied per-resource, only on resources in the dataset's main resource group; a resource without this extra is ignored by the visualization.

Key	Type	Required	Used for
`maille`	string (`fr`\|`region`\|`departement`\|`epci` in this app)	✅	Selects which resource backs a given geo-level in the chart
`value-column`	string	✅	Column in the resource's tabular data holding the measured value
`axes`	object (`{ [axis]: string[] }`)	✅	Declares which columns/values are selectable as chart breakdown axes for this resource

Note: the schema does not constrain maille to an enum; the frontend narrows it to fr | region | departement | epci and maps non-fr values to fixed geocode columns (geocode_region, geocode_departement, geocode_epci) plus a fixed date_mesure year column — those column names are hardcoded conventions, not part of the extras schema.

Data flow for visualization

Dataset extras enable_visualization gates the "Prévisualisation" tab.
Among the dataset's main group resources, only those carrying resource extras ecospheres-indicateurs.maille are candidates.
Selecting a mesh (fr/region/departement/epci) picks the matching resource; selecting a territory (for non-fr meshes) filters rows via the Tabular API using the fixed geocode column for that mesh.
Rows are read from ${tabularApiUrl}/api/resources/{resource.id}/data/, using value-column for the Y value and axes for breakdown/grouping.

Monitoring dashboard

A monitoring dashboard is deployed at indicators-dashboard.sandbox.data.developpement-durable.gouv.fr. It crawls indicator datasets on demo.data.gouv.fr and www.data.gouv.fr and lets the team check, for each one, that its custom metadata (extras) validates against the expected schemas and that its resource files are correctly "APIfied" (reachable and CORS-enabled on the Tabular API) — catching publication issues before they reach the visualization layer.

Github source: https://github.com/abulte/ecospheres-indicators-dashboard