You can read your graph in from a pandas DataFrame and use the connected_component_subgraphs
function (see docs) to split the graph into connected components then and get the largest component from that.
Example reading your graph and making a networkx graph
edge_list_df = pd.read_csv('edges.csv')
g = nx.pandas_edgelist(edge_list_df,source='source',
target='target',edge_attr='weight')
Example getting the connected components and the largest one
component_subgraph_list = list(nx.connected_component_subgraphs(g))
largest_component = max(component_subgraph_list,key=len)
CLICK HERE to find out more related problems solutions.