Geopandas

import geopandas as gpd

import pandas as pd

DataFrame

Datasets

gdf = gpd.read_file("https://naturalearth.s3.amazonaws.com/110m_cultural/ne_110m_admin_0_countries.zip")

 high res land

gdf = gpd.read_file("https://naturalearth.s3.amazonaws.com/10m_physical/ne_10m_land.zip")

IO

US States as a shape file

url = "https://eric.clst.org/assets/wiki/uploads/Stuff/gz_2010_us_040_00_20m.json"

gdf = gpd.read_file(url)

gdf = gdf[gdf.NAME == 'Florida']

Read geoJSON from s3

import geopandas

import s3fs

fs = s3fs.S3FileSystem()

gdf = gpd.read_file(fs.open("s3://BUCKET/file.geojson"))

Read shape from s3

gpd.read_file("zip+s3://BUCKET/file.zip") 

import s3fs

fs = s3fs.S3FileSystem()

with fs.open("s3://BUCKET/file.zip") as f:

    gdf = gpd.read_file(f)

Read parquet

gpd.read_parquet("file.parquet")

gpd.read_parquet(fs.open("file.parquet"))

To parquet

import s3fs

gdf.to_parquet("file.parquet", filesystem=s3fs.S3FileSystem())

To csv

gdf.to_csv("file.csv", index=False)

 To GeoJSON

gdf.to_file("file.geojson", driver="GeoJSON")  

 To shapefile

gdf.to_file("file.shp", driver="GeoJSON")  

points 

gdf = gpd.GeoDataFrame(

    {"geometry": [Point(-83.12, 42.38)]}, geometry="geometry", crs="epsg:4326"

)

df = pd.DataFrame(

    data={

        "time": pd.date_range("2000", freq="D", periods=3),

        "lon": [1, 2, 3],

        "lat": [1, 2, 3],

    }

)

df = pd.DataFrame(

    data={

        "time": pd.date_range("2000-01-04", freq="D", periods=3),

        "lon": [4, 5, 6],

        "lat": [4, 5, 6],

    }

)

gdf = gpd.GeoDataFrame(

    df,

    geometry=gpd.points_from_xy(df["lon"], df["lat"]),

    #geometry=gpd.GeoSeries.from_wkt(df["point_wkt"])

    crs="epsg:4326",

) 

Transform

linestring

from shapely.geometry import LineString

line1 = LineString([(0, 0), (1, 1)]) # (lon, lat)

line2 = LineString([(2, 0), (3, 1)])

gdf = gpd.GeoDataFrame({"geometry": [line1, line2]}, crs="epsg:4326")

geodataframe of points to line string

from shapely.geometry import LineString

LineString(gdf['geometry'].tolist())

geodataframe of line string to points

l = []

for coord in gdf.iloc[0]["geometry"].coords:

    l.append(coord)

_df = pd.DataFrame(l, columns=["lon", "lat"])

_gdf = gpd.GeoDataFrame(

    _df.copy(),

    geometry=gpd.points_from_xy(_df["lon"], _df["lat"]),

    crs="epsg:4326",

).reset_index()

_gdf.plot(aspect=1, column="index")

First point of a linestring

from shapely.geometry import Point

gdf["first_point"] = gdf['geometry'].apply(lambda line: Point(line.coords[0]))

Longitudes from a linestring

Reverse linestrings

from shapely.ops import substring

gdf["geometry"] = gdf["geometry"].apply(lambda x: substring(x, 1, 0, normalized=True))

Meta data

Get bounds of polygon

lon_min, lat_min, lon_max, lat_max = gdf["geometry"].total_bounds

print("lon_min:", lon_min)

print("lat_min:", lat_min)

print("lon_max:", lon_max)

print("lat_max:", lat_max)

Get bounds for rows

gdf["minx"], gdf["miny"], gdf["maxx"], gdf["maxy"] = gdf["geometry"].bounds

Get lon and lat centroid

gdf["latitude"] = gdf.centroid.y

gdf["longitude"] = gdf.centroid.x

Set crs

gdf.set_crs("epsg:4326") # https://epsg.io/4326 (geodic)

Transform crs

gdf.to_crs("EPSG:3395") # https://epsg.io/3395 (mercator)

gdf.to_crs(epsg=3395)

Put multi polygons in one

gdf.dissolve(by='CONTINENT')

Geometry

Get wkt

gdf['wkt'] = gdf["geometry"].to_wkt()

Geocode (city to lon-lat)

See geopy

Indexing

gdf.iloc[[0]] # returns geodataframe

Joins

Field (attribute join)

gdf.merge(df, on="COL")

nearest spatial join

gdf_points.sjoin_nearest(gdf, distance_col="distances_crs")

Smoothing

gdf.simplify(1)

Remove some points 

Plotting

gdf.plot(figsize=(5, 5), edgecolor="purple", facecolor="None")

gdf.plot(column="COL")

gdf.plot(aspect=1)