DataFilterExtension seems to be broken on main (may be on my end though...)

[ATL2001]

Context

While working on the get_category_filter PR, I was getting some strange behavior that seemed to suddenly break what I was working on. I decided to pull main down to my machine and see if possibly something else besides what I was doing somehow caused it. When I run the code below (based on the data filter extension docs, with the colors and slider stripped out) the points from the scatter plot layer aren't rendered on the map.

If I don't add the extension, it renders the points.

If someone could try to recreate this and either confirm that it's not just me, or let me know that it's on my end only I'd really appreciate it. I'm scratching my head right now.

Environment

• OS: Windows 11
• Browser: opera
• Lonboard Version: Main

Steps to reproduce the bug

from pathlib import Path

import geopandas as gpd
import numpy as np
import pandas as pd
import shapely

import lonboard
from lonboard import Map, ScatterplotLayer
from lonboard.layer_extension import DataFilterExtension

url = "https://ookla-open-data.s3.us-west-2.amazonaws.com/parquet/performance/type=mobile/year=2019/quarter=1/2019-01-01_performance_mobile_tiles.parquet"
local_path = Path("data-filter-extension.parquet")
if local_path.exists():
    gdf = gpd.read_parquet(local_path)
else:
    columns = ["avg_d_kbps", "avg_u_kbps", "avg_lat_ms", "devices", "tile"]
    df = pd.read_parquet(url, columns=columns)

    tile_geometries = shapely.from_wkt(df["tile"])
    tile_centroids = shapely.centroid(tile_geometries)
    non_geom_columns = [col for col in columns if col != "tile"]
    gdf = gpd.GeoDataFrame(
        df[non_geom_columns],
        geometry=tile_centroids,
        crs="EPSG:4326",
    )
    gdf.to_parquet(local_path)

filter_extension = DataFilterExtension(filter_size=3)

filter_values = np.column_stack(
    [gdf["avg_d_kbps"], gdf["avg_u_kbps"], gdf["avg_lat_ms"]],
)
initial_filter_range = [
    [10_000, 50_000],
    [1000, 10_000],
    [0, 100],
]

layer = ScatterplotLayer.from_geopandas(
    gdf,
    get_fill_color=[0,0,255],
    radius_units="meters",
    radius_min_pixels=0.1,
    extensions=[filter_extension],
    get_filter_value=filter_values,
    filter_range=initial_filter_range,
)

m = Map(layer, basemap=lonboard.basemap.MaplibreBasemap())

display(m)

[kylebarron]

One initial thing to check is right before the commit bumping to 9.2

[kylebarron]

It is indeed broken. I did a git bisect and 9273319 is the bad commit.

[ATL2001]

oh, thank you! I've been terribly confused as to how/what I did! any ideas on a plan forward?

[kylebarron]

If you want to help debug, you can put a console.log somewhere around here

lonboard/src/model/layer.ts

Lines 827 to 831 in 22e6e7d

	layers.push(
	new GeoArrowScatterplotLayer({
	...this.baseLayerProps(),
	...this.layerProps(batchIdx),
	}),

to see what props are being passed down. It's possible that there's a bug in deck.gl-layers that isn't passing down the filter props correctly

[ATL2001]

it wont be tonight, but I'll make some time to poke around at some point this week

[ATL2001]

I added a log message and everything appears to my eye to make sense from the lonboard side of things:

I've never dug into geoarrow/deck.gl-layers before tonight, but nothing's sticking out to me in there 🙁

[kylebarron]

can you try #977

Screen.Recording.2025-10-26.at.12.35.57.AM.mov

[kylebarron]

Thanks for helping me verify