Ilia Timofeev

``` brew install carlocab/personal/unrar ```` will help on macos

Maybe: ```python alt.Chart(data).mark_point().encode( alt.Color('color:Q', alt.Scale(alt.Scheme(name='viridis', extent=[0, 0.9])) ) ```

@jakevdp for encoding channels it still works in your prototype :). As I understand the goal is to avoid doubling semantics like `scale = alt.Scale()` Chaining methods look very cool...

I know two cases where "types" style wouldn't work: tooltips channel and conditions. For tooltips, we can add an extra virtual channel: `atl.Tooltips("data1:N","data2:T", alt.Tooltip("data_t",timeUnit="day"))` And for conditions add extra parameter...

```python import altair as alt from vega_datasets import data iris = data.iris() alt.Chart(iris).mark_point().encode( x='$(column):Q', y='$(row):Q', color='species' ).repeat('both') ``` JQuery style? or just `x='$column:Q'`

Does the [`CC BY 4.0`](https://creativecommons.org/licenses/by/4.0/) compatible with MIT? I would like to add some examples based on gapminder or world bank data, both sources under `CC BY 4.0` is it...

It's both sides issue. `Geopandas` Antarctica contains point `(180.00000000000014, -90.00000000000003)` which is not valid. And `Vega` do not interpret it as expected. - [x] workaround - [x] issue downstream -...

@nickeubank since https://github.com/altair-viz/altair/pull/1664 Altair has support of geopandas - try it.