def quality(self):
nodes, edges = utils.get_no_redundant_graph()
export_data = {'nodes': [], "links": []}
type2group = {'Actor': 1, 'Message': 2, 'System': 3}
# clustering is needed
# prepare the nodes
for anode in nodes:
export_data['nodes'].append({
'name': anode['text'],
'group': type2group[anode['type']]
})
# prepare the edges
for aedge in edges:
export_data['links'].append({
'source': aedge['source-index'],
'target': aedge['target-index'],
'value': aedge['count']
})
# return the json data to the page visualization
import json
jsonfile = open('ucframe/public/data/ams_matrix.json', 'w')
jsonfile.write(json.dumps(export_data))
jsonfile.close()
return dict(page='quality')
def contribute_pre(self, **kw):
other_input = {}
self_input = {}
self_parameters_name = [
"PreCondition-Status-", "PreCondition-Condition-"
]
self_parameters = []
for i in range(1, 10): #[seLain] TOFIX: No better way to do this ?
for parameters_name in self_parameters_name:
key = parameters_name + str(i)
if key not in self_parameters:
self_parameters.append(key)
remove_key = [] #[seLain] TOFIX: Unknown function code
for i in kw:
if kw[i] == '':
remove_key.append(i) #[seLain] TOFIX: Unknown function code
elif i in self_parameters: #[seLain] TOFIX: Unknown function code
self_input[i] = kw[i]
else: #[seLain] this code seems to be passed
other_input[i] = kw[
i] # to the next page, for params accumulation
# for embedded view test
export_nodes_data, export_edges_data = utils.get_no_redundant_graph()
#return {'parameters':other_input, 'self_input': self_input}
return {
'parameters': other_input,
'self_input': self_input,
'nodes': export_nodes_data,
'edges': export_edges_data
}
def view(self):
export_nodes_data, export_edges_data = utils.get_no_redundant_graph()
# return the json data to the page visualization
import json
jsonfile = open('graph.json', 'w')
jsonfile.write(json.dumps(export_nodes_data))
jsonfile.close()
jsonfile = open('edges.json', 'w')
jsonfile.write(json.dumps(export_edges_data))
jsonfile.close()
return dict(page='view',
nodes=export_nodes_data,
edges=export_edges_data)
def postcon_metrics(self, **kw):
postcon_list = []
for key in kw.keys():
postcon_list.append(key.strip())
current_terms = []
for postcon in postcon_list:
splitted_terms = postcon.split(' ')
for aterm in splitted_terms:
if self.is_ams_term(aterm) and aterm not in current_terms:
current_terms.append(aterm)
print 'current_terms in postcon_metrics', current_terms
# metrics evaluation
nodes, edges = utils.get_no_redundant_graph()
# CRT computation
import metrics
CRT_data = {} ## data to be returned
## calculate the CRT
CRT_dict = metrics.CRT()
CRT_data['terms'] = {}
## get the average CRT
CRT_current_terms = []
for aterm in current_terms:
if aterm in CRT_dict.keys():
CRT_current_terms.append(CRT_dict[aterm])
CRT_data['terms'][aterm] = CRT_dict[
aterm] ## put ony the context terms
if len(CRT_current_terms) == 0:
CRT_average = 0
else:
CRT_average = round(
sum(CRT_current_terms) / len(CRT_current_terms), 2)
CRT_data['average'] = CRT_average
CRT_data['description'] = 'Contributor Ratio of a Term (CRT)' +'\\n\\n' + \
'在目前輸入頁面中您所使用的 AMS 詞彙(Terms), ' + \
'平均來說也被多少比例的其他使用案例貢獻者所使用.' + \
'\\n\\n' + '舉例來說: PuzzleBreaker(57%) 代表有 57% 的貢獻者' + \
'也使用了 PuzzleBreaker 在他們的使用案例中.'
return {'CRT': CRT_data}
def create_tagtable(self, ):
table_location = os.sep.join(
['ucframe', 'model', 'Tag', 'Tag Table.txt'])
TagTable = ConfigObj(table_location, encoding='UTF8')
export_nodes_data, export_edges_data = utils.get_no_redundant_graph()
for anode in export_nodes_data:
TagTable[anode['text']] = {
'Name': anode['text'],
'RealWord': anode['text'],
'Category': anode['type'],
'Type': 'Instance',
'Abstraction': anode['text']
}
TagTable.write()
def cuc_metrics(self, **kw):
event_list = []
for key in kw.keys():
event_list.append(key.strip())
current_terms = []
for event in event_list:
splitted_terms = event.split(' ')
for aterm in splitted_terms:
if self.is_ams_term(aterm) and aterm not in current_terms:
current_terms.append(aterm)
print 'current_terms in cuc_metrics', current_terms
# metrics evaluation
nodes, edges = utils.get_no_redundant_graph()
# calculate the number of senders
# calculate the numder of receivers
'''
NOS = 0
NOR = 0
for aedge in edges:
if current_term == nodes[aedge['target-index']]['text']:
NOS += 1
if current_term == nodes[aedge['source-index']]['text']:
NOR += 1
'''
# CRT computation
import metrics
CRT_data = {} ## data to be returned
## calculate the CRT
CRT_dict = metrics.CRT()
CRT_data['terms'] = {}
## get the average CRT
CRT_current_terms = []
for aterm in current_terms:
if aterm in CRT_dict.keys():
CRT_current_terms.append(CRT_dict[aterm])
CRT_data['terms'][aterm] = CRT_dict[
aterm] ## put ony the context terms
CRT_average = round(sum(CRT_current_terms) / len(CRT_current_terms), 2)
CRT_data['average'] = CRT_average
CRT_data['description'] = 'Contributor Ratio of a Term (CRT)' +'\\n\\n' + \
'在目前輸入頁面中您所使用的 AMS 詞彙(Terms), ' + \
'平均來說也被多少比例的其他使用案例貢獻者所使用.' + \
'\\n\\n' + '舉例來說: PuzzleBreaker(57%) 代表有 57% 的貢獻者' + \
'也使用了 PuzzleBreaker 在他們的使用案例中.'
# LLC computation
NLLC_data = {}
## calculate the LLC
NLLC_result = metrics.NLLC(event_list)
NLLC_data['value'] = NLLC_result
NLLC_data['description'] = 'Normalized Local Logical Conherence (NLLC)' +'\\n\\n' + \
'在目前輸入頁面中您所有敘述之間的邏輯連貫性高低.' + \
'通常越高的 NLLC 代表所有敘述整體越容易被閱讀及理解.' + \
'\\n\\n' + '我們提供了 Average NLLC 做為參考.' + \
'這是其他已經被貢獻的使用案例之平均 NLLC 值.'
ANLLC_result = metrics.ANLLC()
NLLC_data['average'] = ANLLC_result
return {'CRT': CRT_data, 'NLLC': NLLC_data}
def quality_indiv(self):
type_constraint = "Message"
nodes, edges = utils.get_no_redundant_graph()
export_data = {'nodes': [], "links": []}
type2group = {'Actor': 1, 'Message': 2, 'System': 3}
# clustering is needed
# prepare the nodes. only the actor type is considered this time.
for anode in nodes:
if anode['type'] is type_constraint:
export_data['nodes'].append({
'name': anode['text'],
'group': type2group[anode['type']]
})
# prepare the edges
# note that the reverse edges are calculated, not natural from the graph
# each calculated edge is made by comparing the similarity between two nodes
## prepare the node array
## each node array is made of node[linked_node1, linked_node2...]
node_array = {}
for anode in nodes:
if anode[
'type'] is not type_constraint: # only the actor type is considered this time.
continue
node_array[anode['text']] = {}
for aedge in edges:
#print '[in comp]', nodes.index(anode), aedge['source-index']
if nodes.index(anode) is aedge['source-index']:
node_array[anode['text']][
aedge['target-index']] = aedge['count']
## compare similarity
for key_1 in node_array.keys():
for key_2 in node_array.keys():
node_1 = node_array[key_1]
node_2 = node_array[key_2]
# print '[in rev]', node_1, node_2
overlaps = [
min(node_1[x], node_2[x]) for x in node_1 if x in node_2
]
# print '[in rev]', key_1, key_2, overlaps
###### worked !!!!!!!! but the correctness is questionable
###### DO THE CODE REVIEW !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
export_data['links'].append({
'source':
node_array.keys().index(key_1),
'target':
node_array.keys().index(key_2),
'value':
sum(overlaps)
})
'''
for aedge in edges:
export_data['links'].append({'source':aedge['source-index'],
'target':aedge['target-index'],
'value':aedge['count']})
'''
# return the json data to the page visualization
import json
jsonfile = open('ucframe/public/data/ams_matrix_indiv.json', 'w')
jsonfile.write(json.dumps(export_data))
jsonfile.close()
return dict(page='quality_indiv')