11# Explore multiple datasets together
22
3+ ## Current situation
4+
5+ The user experience for exploring datasets currently drives many Jupyter users to open
6+ QGIS or ArcGIS to explore their data, and they report friction switching between
7+ applications.
8+
9+ Some current tools aren't quite meeting user needs, e.g.:
10+
11+ * ` my_gdf.explore() ` : Only shows one dataset at a time
12+ * ` my_dataarray.plot() ` : Shows one dataset, non-interactive
13+ * JupyterGIS: Symbology configuration is manual and low-level
14+ * ` ipyleaflet ` : Low-level. Can't render Xarray DataArrays dynamically (see possible
15+ solution: < https://github.com/davidbrochart/xarray_leaflet > )
16+
17+
18+ ## Proposed improvement
19+
320Prototype: < https://github.com/geojupyter/jupyter-microgis >
421
5- * Add layers to a map for exploration from a Jupyter Notebook:
22+ Add layers to a map for exploration within a Jupyter Notebook:
623
7- ``` python
8- from ... import explore
24+ ``` python
25+ from ... import explore
926
10- explore(
27+ explore(
1128 my_xarray_dataarray,
1229 {' data' : my_geodataframe, ' symbology' : ' choropleth' },
13- )
14- ```
15- * Support rioxarray DataArrays
16- * Support geopandas GeoDataFrames
17- * Maybe: Support WMTS?
30+ )
31+ ```
32+
33+ * Support rioxarray DataArrays
34+ * Support geopandas GeoDataFrames
35+ * Maybe: Support WMTS?
1836* Support some curated default symbology options
1937 * Choropleth: number of steps, classification mode, ?
2038 * Symbol map: shape, min/max size, ...
@@ -34,3 +52,12 @@ Prototype: <https://github.com/geojupyter/jupyter-microgis>
3452 * Plain language search
3553 * Produce Python one-liners to bring that dataset into their notebook, e.g. ` geopandas.read_file(...) ` and ` xarray.open_mfdataset(...) `
3654* Display the data on slippy map widget (e.g. DeckGL)
55+
56+
57+ ## Questions
58+
59+ ### Is this ` leafmap ` ?
60+
61+ Is this idea basically [ ` leafmap ` ] ( https://leafmap.org/ ) ?
62+ If we add support to directly visualize xarray datasets with TiTiler in leafmap, does
63+ that meet the goals of this use case?
0 commit comments