updates for geo (longitude/latitude) #799
Conversation
…ctions to add kelper configurations to graphistry plottable
|
note: mercator layout docs need improvment |
| # Create geographic visualization | ||
| g = graphistry.edges(edges_df, 'src', 'dst').nodes(nodes_df, 'nodeId') | ||
| g = g.bind(point_longitude='lon', point_latitude='lat') | ||
| g.plot() |
There was a problem hiding this comment.
doesn't this need .settings(url_params={'play': 0}) ?
|
|
||
| Visualize nodes as points on a map: | ||
|
|
||
| .. code-block:: python |
There was a problem hiding this comment.
What data does this match up to? The above example has nodeId but no col Cities... this example feels incomplete->confusing...
| .. code-block:: python | ||
| # Configure dataset and arc layer for edges | ||
| g = (g |
There was a problem hiding this comment.
same... maybe include the edge table?
| "metadata": {}, | ||
| "outputs": [], | ||
| "source": [ | ||
| "g.plot()\n" |
There was a problem hiding this comment.
save ipynb run against hub with embedded public viz in these
| " \"type\": random_types\n", | ||
| "})\n", | ||
| "\n", | ||
| "display(edges_df.head(3))" |
There was a problem hiding this comment.
also let's show key df outputs like this as explains the following data
| "outputs": [], | ||
| "source": [ | ||
| "# Create Kepler encoding with datasets and layers\n", | ||
| "g2 = g \\\n", |
There was a problem hiding this comment.
Is there a 'minimal' simple map we can get generated 'for free' on this data ? requiring all this config for a maps hello world on it looks quite complicated...
There was a problem hiding this comment.
ex: should the first example be mercator? does that trigger kepler?
| @@ -0,0 +1,841 @@ | |||
| """ | |||
There was a problem hiding this comment.
can we move into something like https://github.com/graphistry/pygraphistry/tree/master/graphistry/plugins ?
lmeyerov
left a comment
lmeyerov
left a comment
There was a problem hiding this comment.
See comments -
- file locations
- saving outputs for iynb versions, both viz + df
- 'minimal' kepler example looks scary.. is that inherent / api issue needing simplification, or demo can provide an easier minimal starting example for that data?
- i'm unclear on diff of mercator layout vs kepler encoding... maybe discuss in both & cross-link?
point_longitudeandpoint_latitudefor specifying columns containing geographic coordinatesencode_kepler_dataset()- Add datasets (nodes, edges, countries, states, etc.)encode_kepler_layer()- Add visualization layers (point, arc, line, grid, hexagon, etc.)encode_kepler()- Apply complete Kepler configurationmercator_layout()- Convert lat/lon to Mercator projection with GPU/CPU supportKeplerDataset,KeplerLayer,KeplerEncodingclasses for type-safe configurationtest_kepler.pyand 6 tests intest_layout.py