Reputation: 121
EDIT : Adding code that generates the polygons. df
is the base data loaded from csv and processed into clusters :
df.head(1)
Requirement PickupTime Year Month DayOfWeek Hour Town StreetName StreetNr Latitude Longitude GeoHash cluster dbcluster
0 NaN 2013-01-21 13:20:17.000 2013 1 1 13 Mikkeli PORRASSALMENKATU 16 61.68844 27.274791 udgquw 17 29
Then create polygons by combining points that fall into a cluster :
polygons = []
for i in range(30):
latlong = df[['Latitude','Longitude']].loc[df['cluster'] == i]
polygons.append([i,MultiPoint(latlong.to_numpy()).convex_hull
,centerOfRegions[i][0],centerOfRegions[i][1]])
Then load this into a dataframe
polygondf = pd.DataFrame(polygons, columns=['cluster', 'geometry','center_lat','center_long'])
#polygondf['count'] = 1
polygondf.head(2)
cluster geometry center_lat center_long
0 0 POLYGON ((61.6947869 27.2714365, 61.6943701 27... 61.692932 27.278602
1 1 POLYGON ((61.699156 27.2833976, 61.69188 27.28... 61.696324 27.290567
Now load this dataframe object into a GeoDataFrame object :
import geopandas as gpd
frm = gpd.GeoDataFrame(polygondf, crs="EPSG:4326", geometry='geometry')
frm.head()
Output:
cluster geometry center_lat center_long count
0 0 POLYGON ((61.69479 27.27144, 61.69437 27.27188... 61.692932 27.278602 1
1 1 POLYGON ((61.69916 27.28340, 61.69188 27.28595... 61.696324 27.290567 1
2 2 POLYGON ((62.02711 27.76096, 61.90189 27.79479... 61.843740 28.136880 1
3 3 POLYGON ((61.81762 27.06707, 61.56495 27.11654... 61.681207 27.183761 1
4 4 POLYGON ((61.69780 27.25109, 61.68140 27.25391... 61.689157 27.259389 1
The polygons in frm['geometry']
are valid and can draw a plot as follows:
frm.plot(figsize=(6, 6))
plt.show()
But when try to draw the same polygons on a folium map, the map shows no polygons drawn.
m = folium.Map([61.681659, 27.271036], zoom_start=8, tiles='cartodbpositron')
for _, r in frm.iterrows():
sim_geo = gpd.GeoSeries(r['geometry'],crs="EPSG:4326").simplify(tolerance=0.001)
#sim_geo = sim_geo.to_crs("EPSG:4326")
geo_j = sim_geo.to_json()
geo_j = folium.GeoJson(data=geo_j,
style_function=lambda x: {'fillColor': 'orange'})
folium.Popup(r['cluster']).add_to(geo_j)
geo_j.add_to(m)
m
Can I know what could cause this issue? Folium maps is working fine and I have been able to draw polygons as a Choropleth as well. But this is something I am unable to fix.
Upvotes: 3
Views: 949
Reputation: 799
frm
probably does not have the correct crs to be show on the folium map.
Even if creating a GeoDataFrame with the crs parameter it seems to not change the original crs.
polygondf = gpd.read_file('shapefile.shp')
frm = gpd.GeoDataFrame(polygondf, crs="EPSG:4326", geometry='geometry')
frm.crs # epsg:3006
frm = frm.to_crs(epsg=4326) # now crs is updated
EDIT:
The shapely MultiPoint has a constructor of (x, y[, z ])
which would mean order of (longitude, latitude), see this post. Create your polgyons in longlat instead of latlong order.
Upvotes: 2