Use position-based index to avoid reaggregation issues in lines_polys_handler

[luseverin]

Changes proposed in this PR:

• User-inputted exposures indices are ignored during re-aggregation phase of the lines_polys_handler module

This PR fixes #992

Hi!
As discussed in #992 , I tried to implement a quick fix to avoid erreoneous re-aggregation of disaggregated exposures to their original shape which was caused by non-unique index in the exposures GeoDataFrame. To fix it, I simply replaced the original indices by positional indices in the csr_mask used for the impact_mat reaggregation in _aggregate_impact_mat. This fix seems to do the job for this specific case but I am not entirely sure it works as intended as I was a bit confused about what was going on in the _aggregate_impact_mat function. For instance, why did we use the original index values to make the csr_mask as anyways the csr matrices work with positional indices to my understanding. Also, @peanutfun, you can let me know if this fix corresponds to the best solution you mentionned in #992 or not.

As a reminder, this is the code that used to create the error (showing up when plotting):

from shapely.geometry import Point, LineString, Polygon
import geopandas as gpd
gdf_test = gpd.GeoDataFrame
data = {
    "name": ["A", "B", "C"],
    "value": [10, 20, 30],
    "geometry": [
        Point(4, 51),            
        Point(4, 52),
        Point(4, 53)
    ],
}
gdf_test = gpd.GeoDataFrame(data, crs="EPSG:4326") 
gdf_test.index = [0,1,0] #need to specify non-unique index to cause the error!

#hazard and impf
from climada.util.api_client import Client
import climada.util.lines_polys_handler as u_lp
from climada.entity.impact_funcs import ImpactFuncSet
from climada.entity.impact_funcs.storm_europe import ImpfStormEurope
from climada.entity import Exposures
from climada.engine import ImpactCalc


haz = Client().get_hazard("storm_europe", name="test_haz_WS_nl", status="test_dataset")

#exposures
exp_line = Exposures(gdf_test)

#impact function
impf_line = ImpfStormEurope.from_welker()
impf_set = ImpactFuncSet([impf_line])

exp_line.data["impf_WS"] = 1  # specify impact function

#Break down calc_geom_impact
# disaggregate exposure
exp_pnt = u_lp.exp_geom_to_pnt(
    exp=exp_line,
    res=500,
    to_meters=True,
    disagg_met=u_lp.DisaggMethod.FIX,
    disagg_val=1e5,
)
exp_pnt.assign_centroids(haz)

# compute point impact
calc = ImpactCalc(exp_pnt, impf_set, haz)
impact_pnt = calc.impact(save_mat=True, assign_centroids=False)

# re-aggregate impact to original exposure geometry
impact_agg = u_lp.impact_pnt_agg(impact_pnt, exp_pnt.gdf, u_lp.AggMethod.SUM)

# plot
u_lp.plot_eai_exp_geom(impact_agg);

PR Author Checklist

• Read the Contribution Guide
• Correct target branch selected (if unsure, select develop)
• Descriptive pull request title added
• Source branch up-to-date with target branch
• Documentation updated
• Tests updated
• Tests passing
• No new linter issues
• Changelog updated

PR Reviewer Checklist

• Read the Contribution Guide
• CLIMADA Reviewer Checklist passed
• Tests passing
• No new linter issues