def example_partition(): df = gp.read_file(os.path.join(TEST_DATA_PATH, "mo_cleaned_vtds.shp")) with open(os.path.join(TEST_DATA_PATH, "MO_graph.json")) as f: graph_json = json.load(f) graph = networkx.readwrite.json_graph.adjacency_graph(graph_json) assignment = get_assignment_dict(df, "GEOID10", "CD") add_data_to_graph( df, graph, ['PR_DV08', 'PR_RV08', 'POP100', 'ALAND10', 'COUNTYFP10'], id_col='GEOID10') updaters = { **votes_updaters(['PR_DV08', 'PR_RV08'], election_name='08'), 'population': Tally('POP100', alias='population'), 'counties': county_splits('counties', 'COUNTYFP10'), 'cut_edges': cut_edges, 'cut_edges_by_part': cut_edges_by_part } return Partition(graph, assignment, updaters)
def PA_partition(): # this is a networkx adjancency data json file with CD, area, population, and vote data graph = construct_graph("./testData/PA_graph_with_data.json") # Add frozen attributes to graph # data = gp.read_file("./testData/frozen.shp") # add_data_to_graph(data, graph, ['Frozen'], 'wes_id') assignment = dict( zip(graph.nodes(), [graph.node[x]['CD'] for x in graph.nodes()])) updaters = { **votes_updaters(['VoteA', 'VoteB']), 'population': Tally('POP100', alias='population'), 'perimeters': perimeters, 'exterior_boundaries': exterior_boundaries, 'boundary_nodes': boundary_nodes, 'cut_edges': cut_edges, 'areas': Tally('ALAND10', alias='areas'), 'polsby_popper': polsby_popper, 'cut_edges_by_part': cut_edges_by_part } return Partition(graph, assignment, updaters)
def setup_for_proportion_updaters(columns): graph = three_by_three_grid() attach_random_data(graph, columns) assignment = random_assignment(graph, 3) updaters = votes_updaters(columns) return Partition(graph, assignment, updaters)
def test_vote_proportion_returns_nan_if_total_votes_is_zero(): columns = ['D', 'R'] graph = three_by_three_grid() for node in graph.nodes: for col in columns: graph.nodes[node][col] = 0 updaters = votes_updaters(columns) assignment = random_assignment(graph, 3) partition = Partition(graph, assignment, updaters) assert all(math.isnan(value) for value in partition['D%'].values()) assert all(math.isnan(value) for value in partition['R%'].values())
def set_up_chain(plan, total_steps, adjacency_type='queen'): graph = Graph.load(f"./PA_{adjacency_type}.json").graph assignment = {node: graph.nodes[node][plan] for node in graph.nodes} updaters = { **votes_updaters(elections["2016_Presidential"], election_name="2016_Presidential"), **votes_updaters(elections["2016_Senate"], election_name="2016_Senate"), 'population': Tally('population', alias='population'), 'perimeters': perimeters, 'exterior_boundaries': exterior_boundaries, 'interior_boundaries': interior_boundaries, 'boundary_nodes': boundary_nodes, 'cut_edges': cut_edges, 'areas': Tally('area', alias='areas'), 'polsby_popper': polsby_popper, 'cut_edges_by_part': cut_edges_by_part } partition = Partition(graph, assignment, updaters) population_constraint = within_percent_of_ideal_population(partition, 0.01) compactness_constraint = SelfConfiguringLowerBound(L_minus_1_polsby_popper, epsilon=0.1) is_valid = Validator(default_constraints + [population_constraint, compactness_constraint]) return partition, MarkovChain(propose_random_flip, is_valid, always_accept, partition, total_steps)
def test_vote_proportion_updater_returns_percentage_or_nan_on_later_steps(): columns = ['D', 'R'] graph = three_by_three_grid() attach_random_data(graph, columns) assignment = random_assignment(graph, 3) updaters = {**votes_updaters(columns), 'cut_edges': cut_edges} initial_partition = Partition(graph, assignment, updaters) chain = MarkovChain(propose_random_flip, Validator([no_vanishing_districts]), lambda x: True, initial_partition, total_steps=10) for partition in chain: assert all(is_percentage_or_nan(value) for value in partition['D%'].values()) assert all(is_percentage_or_nan(value) for value in partition['R%'].values())
def example_partition(): df = gp.read_file("./testData/mo_cleaned_vtds.shp") with open("./testData/MO_graph.json") as f: graph_json = json.load(f) graph = networkx.readwrite.json_graph.adjacency_graph(graph_json) assignment = get_assignment_dict(df, "GEOID10", "CD") add_data_to_graph( df, graph, ['PR_DV08', 'PR_RV08', 'POP100', 'ALAND10', 'COUNTYFP10'], id_col='GEOID10') updaters = { **votes_updaters(['PR_DV08', 'PR_RV08'], election_name='08'), 'population': Tally('POP100', alias='population'), 'areas': Tally('ALAND10', alias='areas'), 'counties': county_splits('counties', 'COUNTYFP10'), 'perimeters': perimeters, 'exterior_boundaries': exterior_boundaries, 'boundary_nodes': boundary_nodes, 'polsby_popper': polsby_popper, 'cut_edges': cut_edges, 'cut_edges_by_part': cut_edges_by_part } return Partition(graph, assignment, updaters)
'perimeters': perimeters, 'exterior_boundaries': exterior_boundaries, 'interior_boundaries': interior_boundaries, 'boundary_nodes': boundary_nodes, 'cut_edges': cut_edges, 'areas': Tally('areas'), 'polsby_popper': polsby_popper, 'cut_edges_by_part': cut_edges_by_part, #'County_Splits': county_splits('County_Splits',county_col) } # Add the vote updaters for multiple plans for i in range(num_elections): updaters = { **updaters, **votes_updaters(election_columns[i], election_names[i]) } # This builds the partition object initial_partition = Partition(graph, assignment, updaters) # Choose which binary constraints to enforce # Options are in validity.py pop_limit = .2 population_constraint = within_percent_of_ideal_population( initial_partition, pop_limit) compactness_constraint_Lm1 = LowerBound( L_minus_1_polsby_popper, L_minus_1_polsby_popper(initial_partition))
add_data_to_graph(df, graph, data_list, id_col=unique_label) # Desired proposal method proposal_method = propose_random_flip_no_loops # Desired proposal method acceptance_method = always_accept # Number of steps to run steps = 1000 print("loaded data") # Necessary updaters go here updaters = { **votes_updaters([vote_col1, vote_col2]), 'population': Tally(pop_col, alias='population'), 'perimeters': perimeters, 'exterior_boundaries': exterior_boundaries, 'interior_boundaries': interior_boundaries, 'boundary_nodes': boundary_nodes, 'cut_edges': cut_edges, 'areas': Tally(area_col, alias='areas'), 'polsby_popper': polsby_popper,
def read_basic_config(configFileName): """Reads basic configuration file and sets up a chain run :configFileName: relative path to config file :returns: Partition instance and MarkovChain instance """ # set up the config file parser config = configparser.ConfigParser() config.read(configFileName) # SET UP GRAPH AND PARTITION SECTION # create graph and get global names for required graph attributes graph, POP, AREA, CD = gsource_gdata(config, 'GRAPH_SOURCE', 'GRAPH_DATA') voteDataList = vsource_vdata(graph, config, 'VOTE_DATA_SOURCE', 'VOTE_DATA') # create a list of vote columns to update DataUpdaters = {**votes_updaters(voteDataList)} # Previously used individual columns # {v: updates.Tally(v) for v in voteDataList} # original plan for fixing %'s: # for v in voteDataList: # DataUpdaters = {**DataUpdaters, v+"%": updates.Tally(v+"%")} # construct initial districting plan assignment = {x[0]: x[1][CD] for x in graph.nodes(data=True)} # set up validator functions and create Validator class instance validatorsUpdaters = [] validators = [] if config.has_section('VALIDITY') and len(list( config['VALIDITY'].keys())) > 0: validators = list(config['VALIDITY'].values()) for i, x in enumerate(validators): if len(x.split(',')) == 1: validators[i] = getattr(valids, x) else: [y, z] = x.split(',') validators[i] = valids.WithinPercentRangeOfBounds( getattr(valids, y), z) validatorsUpdaters.extend( [x.split(',')[0] for x in config['VALIDITY'].values()]) validators = valids.Validator(validators) # add updaters required by this list of validators to list of updaters for x in validatorsUpdaters: DataUpdaters.update(dependencies(x, POP, AREA)) # END SET UP GRAPH AND PARTITION SECTION # SET UP MARKOVCHAIN RUN SECTION # set up parameters for markovchain run chainparams = config['MARKOV_CHAIN'] # number of steps to run num_steps = 1000 if 'num_steps' in list(chainparams.keys()): num_steps = int(chainparams['num_steps']) # type of flip to use proposal = proposals.propose_random_flip if 'proposal' in list(chainparams.keys()): proposal = getattr(proposals, chainparams['proposal']) # acceptance function to use accept = accepts.always_accept if 'accept' in list(chainparams.keys()): accept = getattr(accepts, chainparams['accept']) # END SET UP MARKOVCHAIN RUN SECTION # SET UP DATA PROCESSOR FOR CHAIN RUN # get evaluation scores to compute and the columns to use for each escores, cfunc, elist, sVisType, outFName = escores_edata( config, "EVALUATION_SCORES", "EVALUATION_SCORES_DATA") # add evaluation scores updaters to list of updators for x in elist: DataUpdaters.update(dependencies(x, POP, AREA)) # END SET UP DATA PROCESSOR FOR CHAIN RUN updaters = DataUpdaters # create markovchain instance initial_partition = Partition(graph, assignment, updaters) chain = MarkovChain(proposal, validators, accept, initial_partition, num_steps) return chain, cfunc, escores, sVisType, outFName
def main(): #graph = construct_graph_from_file("/Users/caranix/Desktop/Alaska_Chain/AK_data.shp", geoid_col="DISTRICT") with open('./alaska_graph.json') as f: data = json.load(f) graph = networkx.readwrite.json_graph.adjacency_graph(data) df = gp.read_file( "/Users/caranix/Desktop/Alaska_Chain/AK_data.shp" ) # assignment = dict(zip(graph.nodes(), [graph.node[x]['HOUSEDIST'] for x in graph.nodes()])) add_data_to_graph(df, graph, [ 'join_Distr', 'POPULATION', 'join_Dem', 'join_Rep', 'perc_Dem', 'perc_Rep', 'AREA' ], id_col='DISTRICT') data = json.dumps(networkx.readwrite.json_graph.adjacency_data(graph)) with open('./alaska_graph.json', 'w') as f: f.write(data) assignment = dict( zip(graph.nodes(), [graph.node[x]['join_Distr'] for x in graph.nodes()])) updaters = { 'population': Tally('POPULATION', alias='population'), 'cut_edges': cut_edges, 'cut_edges_by_part': cut_edges_by_part, **votes_updaters(['join_Dem', 'join_Rep'], election_name='12'), 'perimeters': perimeters, 'exterior_boundaries': exterior_boundaries, 'boundary_nodes': boundary_nodes, 'cut_edges': cut_edges, 'areas': Tally('AREA', alias='areas'), 'polsby_popper': polsby_popper } p = Partition(graph, assignment, updaters) print("Starting Chain") chain = BasicChain(p, 1000000) allAssignments = {0: chain.state.assignment} for step in chain: allAssignments[chain.counter + 1] = step.flips # print(mean_median(step, 'join_Dem%')) # with open("chain_outputnew.json", "w") as f: # f.write(json.dumps(allAssignments)) #efficiency_gap(p) # mean_median(p, 'join_Dem%') scores = { 'Mean-Median': functools.partial(mean_median, proportion_column_name='join_Dem%'), 'Mean-Thirdian': functools.partial(mean_thirdian, proportion_column_name='join_Dem%'), 'Efficiency Gap': functools.partial(efficiency_gap, col1='join_Dem', col2='join_Rep'), 'L1 Reciprocal Polsby-Popper': L1_reciprocal_polsby_popper } initial_scores = {key: score(p) for key, score in scores.items()} table = pipe_to_table(chain, scores) fig, axes = plt.subplots(2, 2) quadrants = { 'Mean-Median': (0, 0), 'Mean-Thirdian': (0, 1), 'Efficiency Gap': (1, 0), 'L1 Reciprocal Polsby-Popper': (1, 1) } for key in scores: quadrant = quadrants[key] axes[quadrant].hist(table[key], bins=50) axes[quadrant].set_title(key) axes[quadrant].axvline(x=initial_scores[key], color='r') plt.show() '''