I created a multipolygon (shapefile) from a PNG image. The PNG image was originally a GeoTIFF. The geometries of my multipolygon are values of a normal x,y axis (with 0,0 as origin). In other words, the geometries are not lat,long. See sample:
geometry
0 POLYGON ((0.000 0.000, 0.000 310.000, 5.000 31...
1 POLYGON ((160.000 0.000, 160.000 3.000, 159.00...
2 POLYGON ((794.000 219.000, 794.000 0.000, 443....
3 POLYGON ((555.000 451.000, 793.000 451.000, 79...
4 POLYGON ((417.000 451.000, 555.000 451.000, 55...
5 POLYGON ((237.000 196.000, 237.000 205.000, 23...
I want to project my polygon in a specific CRS and location on Earth. This polygon should match the bbox and CRS of the original GeoTIFF. My multipolygon has no CRS defined.
How can I project my multipolygon? I tried the following, and got the error: AttributeError: 'GeoDataFrame' object has no attribute 'set_crs'
import geopandas as gpd
fp = r"C:\foo\intersection_all2.shp"
data = gpd.read_file(fp)
data['geometry'].head()
# Let's make a copy of our data
orig = data.copy()
# Reproject the data
data = data.set_crs(epsg=32618)
I am searching a solution for Python 3.7, Windows 10.
it says no attribute, so set crs at the object property like this
gdf.crs = {"init":"epsg:4326"}
then you can project it to another by to_crs(), for example
gdf.to_crs(epsg=3857)