出発地点が、時間帯によって、vendor間の違いがあるか、グラフ化して見る。https://colab.research.google.com/drive/1huNXbUTTg57sJf7jLb1Rs00yUR3WHb75?usp=sharing

pointsdistbear.py

import marimo

__generated_with = "0.16.5"
app = marimo.App(width="full")


@app.cell
def _(mo):
    mo.md(
        r"""
        ## 出発地点が、時間帯によって、vendor間の違いがあるか、グラフ化して見る
    
        https://colab.research.google.com/drive/1huNXbUTTg57sJf7jLb1Rs00yUR3WHb75?usp=sharing
        """
    )
    return


@app.cell
def _():
    import marimo as mo
    return (mo,)


@app.cell
def _():
    import geopandas as gpd
    import pandas as pd

    EPSG= "WGS84"

    df = pd.read_json("https://github.com/visgl/deck.gl-data/raw/refs/heads/master/examples/trips/trips-v7.json")
    df["lonlat"] = df["path"].apply(lambda path: path[0])
    df["lon"] = df["lonlat"].apply(lambda lonlat: lonlat[0])
    df["lat"] = df["lonlat"].apply(lambda lonlat: lonlat[1])
    df["tm"] = df["timestamps"].apply(lambda timestamps: int(timestamps[0] / 100))
    df = df.drop(columns=["timestamps", "path", "lonlat"])
    gdf = gpd.GeoDataFrame(df, geometry=gpd.points_from_xy(df["lon"], df["lat"])).set_crs(EPSG)
    return EPSG, gdf, gpd


@app.cell
def _(gdf):
    import plotly.express as px
    figlon = px.box(gdf.drop(columns=["geometry"]), x="tm", y="lon", color="vendor")
    figlon.show()
    figlat = px.box(gdf.drop(columns=["geometry"]), x="tm", y="lat", color="vendor")
    figlat.show()
    return (px,)


@app.cell
def _(mo):
    mo.md(r"### 重心地点からの距離と方位角をグラフ化してみる")
    return


@app.cell
def _(EPSG, gdf, gpd):
    # Centroid for points
    # cf. https://stackoverflow.com/a/70088741
    import shapely
    utmcrs = gdf.estimate_utm_crs()
    gdfutm = gdf.to_crs(utmcrs)

    mp = shapely.geometry.MultiPoint(gdfutm.geometry)
    centroidutm = mp.centroid
    centroid = gpd.GeoSeries([centroidutm], crs=utmcrs).to_crs(EPSG)[0]

    # centroid = gdfutm.geometry.union_all().centroid
    # print(gpd.GeoSeries([centroid], crs=utmcrs).to_crs(EPSG))
    return (centroid,)


@app.cell
def _(centroid, gdf):
    import pyproj
    geod = pyproj.Geod(ellps="WGS84")
    gdf["bearing"], _, gdf["distance"] = geod.inv([centroid.x] * len(gdf), [centroid.y] * len(gdf), gdf.geometry.x, gdf.geometry.y)
    #gdf.drop(columns=["lon", "lat"], inplace=True)
    return


@app.cell
def _(gdf):
    gdf.groupby("vendor").describe()
    return


@app.cell
def _(gdf, px):
    fig = px.box(gdf.drop(columns=["geometry"]), x="tm", y="bearing", color="vendor")
    fig.show()
    return


@app.cell
def _(gdf, px):
    fig2 = px.box(gdf.drop(columns=["geometry"]), x="tm", y="distance", color="vendor")
    fig2.show()
    return


if __name__ == "__main__":
    app.run()