Python: how to create square buffers around points with a distance in meters

Question

I have a geopandas dataframe gdf

gdf
    ID  longitude   latitude    geometry
0   80  103.619501  1.2810      POINT (103.619500987 1.281)
1   81  103.619501  1.2855      POINT (103.619500987 1.2855)

Following this suggestion I create a square buffer around it with a distance bd defined as:

bd = abs((gdf['latitude'][0]-gdf['latitude'][1])/2)

And finally I am able to get the following:

buffer = gdf.buffer(bd)
envelope = buffer.envelope

f, ax = plt.subplots(figsize=(7.5, 7.5))
envelope.plot(color='white', edgecolor='gray',ax=ax)
gdf.plot(ax=ax)

How can I set a distance bd that corresponds to 500 meters?

https://stackoverflow.com/questions/57284066/calculate-geographical-distance-between-5-cities-with-all-the-possible-combinati/57287599#57287599 Does this link help? — ComplicatedPhenomenon, Aug 15 '19 at 09:52

martinfleis · Answer 1 · 2021-03-20T15:24:04.853

7

You need to reproject your data to CRS which uses meters as coordinates. As you are using longitude and latitude, your values are in degrees. All, including buffer distance. If you want to do square buffer, you do not need to use envelope, just set cap_style to 3 (see shapely docs).

gdf.crs = 'epsg:4326' # I am assuming here
gdf = gdf.to_crs(epsg=3395)

buffer = gdf.buffer(500, cap_style=3) # you might want to use 250, guessing from your image

For more details I recommend geopandas user guide.

edited Mar 20 '21 at 15:24

answered Aug 15 '19 at 10:47

martinfleis

7,124
2
22
30

Do you loss the attribute table information after doing the buffering of points? – Roger Almengor Feb 15 '21 at 06:49
Not if you assign the GeoSeries containing buffer as a column. Either to replace original `geometry` column or a new one. – martinfleis Feb 15 '21 at 08:47

score 2 · Answer 2 · edited Dec 03 '20 at 14:00

2

this is my function, which have high precision. data is a DataFrame,which contain a shapely's geometry type

def generate_buffer_meter(data, radiu, geometry='geometry', crs='epsg:4326'):
    data = gpd.GeoDataFrame(data, geometry=geometry, crs=crs)
    data = data.to_crs('+proj=aeqd +units=m  +x_0=0 +y_0=0')
    data[geometry] = data[geometry].buffer(radiu)
    data = data.to_crs(crs)
    return data

edited Dec 03 '20 at 14:00

Boken

4,825
10
32
42

answered Dec 03 '20 at 13:35

happy hacking

21
1

your method returns circle, not square. You may use `cap_style=3` as in answer above – VB_ Jul 05 '21 at 23:18
I think you want .to_crs(epsg=3395) instead of aeqd. And, .buffer(radiu,cap_style=3) – jonincanada Jun 09 '22 at 16:12

Python: how to create square buffers around points with a distance in meters

2 Answers2