pandapower.select_subnet() function changes the dtype of the bus index

[robertobarrio98]

It changes the bus indexes dtype from 'int64' to 'uint64'. With this change, if you want to pass that subnet to a networkx graph with impedances added as weights, using:

topology.create_nxgraph(subnet, calc_branch_impedances=True)

an error will raise: "TypeError: 'numpy.float64' object cannot be interpreted as an integer"

Tried changing manually the type to int64 with subnet.bus.index = subnet.bus.index.astype('int64') and then it worked fine

[SteffenMeinecke]

Dear @robertobarrio98 could you please provide a minimal example where this problem occur? I cannot reproduce it. The following code

import pandapower as pp
import pandapower.networks as pn

net = pn.case9()
subnet = pp.select_subnet(net, [0, 3, 4, 5])

print(type(subnet.bus.index[0]))
print(subnet.bus.index.dtype)

leads to

<class 'numpy.int64'>
int64

on my machine (with pp v2.10.1 and v2.11.1 as well).