def test_get_vertices_values(self):
"""Tests get_vertices_values."""
test1 = io.StringIO('1\t2\t0.1\n2\t3\t0.2\n1\t3\t0.3')
expected1 = [('1', '2', 0.1), ('2', '3', 0.2), ('1', '3', 0.3)]
actual1 = list(edge_analysis.get_vertices_values(test1))
self.assertEqual(actual1, expected1)
test2 = io.StringIO('')
expected2 = []
actual2 = list(edge_analysis.get_vertices_values(test2))
self.assertEqual(actual2, expected2)
def test_get_vertices_values(self):
"""Tests get_vertices_values."""
test1 = io.StringIO('1\t2\t0.1\n2\t3\t0.2\n1\t3\t0.3')
expected1 = [('1', '2', 0.1), ('2', '3', 0.2), ('1', '3', 0.3)]
actual1 = list(edge_analysis.get_vertices_values(test1))
self.assertEqual(actual1, expected1)
test2 = io.StringIO('')
expected2 = []
actual2 = list(edge_analysis.get_vertices_values(test2))
self.assertEqual(actual2, expected2)
def index(field, weightedness):
"""Renders chord diagram for a specific field.
Args:
field (string): A field, e.g. "major1"
weightedness (string): Either weighted, unweighted, or weighted_norm
"""
network_filename = '../data/reduced/network_{}_{}.tsv'.format(
field, weightedness)
try:
with open(network_filename, 'r') as network_file:
full_network = list(edge_analysis.get_vertices_values(network_file))
except FileNotFoundError:
bottle.abort(404, 'Couldn\'t find correct data file!')
rendered_network = render_network_to_json(full_network, field)
return bottle.template('chord', field=field, data=rendered_network)
def index(field, weightedness):
"""Renders chord diagram for a specific field.
Args:
field (string): A field, e.g. "major1"
weightedness (string): Either weighted, unweighted, or weighted_norm
"""
network_filename = '../data/reduced/network_{}_{}.tsv'.format(
field, weightedness)
try:
with open(network_filename, 'r') as network_file:
full_network = list(
edge_analysis.get_vertices_values(network_file))
except FileNotFoundError:
bottle.abort(404, 'Couldn\'t find correct data file!')
rendered_network = render_network_to_json(full_network, field)
return bottle.template('chord', field=field, data=rendered_network)
help='Either "weighted" or "unweighted".')
parser.add_argument(
'file', nargs='?', type=argparse.FileType('r'),
default=sys.stdin, help='File to segment (default standard input).')
parser.add_argument(
'--student-archive-path', dest='student_archive_path',
help='Path of the student archive file.', required=True)
parser.add_argument(
'--swig-module-path', dest='swig_module_path',
help='Directory containing the swig modules.',
action=ReadableDirectory, required=True)
args = parser.parse_args()
# read edges from file; these correspond to actual interaction values
edge_lines = list(edge_analysis.get_vertices_values(args.file))
# get expected interaction values
students_wrapper = StudentContainerWrapper(
args.swig_module_path, args.student_archive_path)
segment_counts = segment.get_segment_counts(students_wrapper, args.field)
# compare actual edges with expected edges, put into matrix
segments = sorted(list(segment_counts.keys()))
population = sum(segment_counts.values())
heatmatrix = numpy.zeros((len(segments), len(segments)))
for vertex1, vertex2, actual_weight in edge_lines:
vertex1_weight = segment_counts[vertex1]
vertex2_weight = segment_counts[vertex2]
# store vertex expected weight
import numpy
import edge_analysis
import simpson_diversity
if __name__ == '__main__':
parser = argparse.ArgumentParser(
description='Calculate simpson diversity index for every vertex in '
'a network.')
parser.add_argument(
'file', nargs='?', type=argparse.FileType('r'),
default=sys.stdin, help='File to segment (default standard input).')
args = parser.parse_args()
edge_lines = edge_analysis.get_vertices_values(args.file)
edge_values = edge_analysis.reduce_to_vertices(edge_lines)
for vertex, values in edge_values.items():
# make every value in values an array because calculation functions
# expect multiple arrays grouped by segment
values_sum = numpy.sum(values)
single_value_arrays = (numpy.array(value) for value in values)
# We use weighted regardless of the input file because we the value to
# be taken into account. There is only one value, so looking at the
# number of values is useless.
diversity_index = (
simpson_diversity.calculate_weighted_simpson_diversity_index(
single_value_arrays, values_sum))