在校学生一枚，需要用一段 调用jar包的python程序，无奈没有python基础，请各位帮忙看下，教教我如何使用？因为包内没给任何使用说明文件，主模块代码也就200行左右，我贴到下面，还请帮忙：
请教这个包如何使用，该怎样运行起来，模型可以输入输出，里面的函数能赋值运行就行。
包内是调用了一个jar包，通过cmd命令行运行，这个我我已经会自行使用了，但是通过这个python调用的方法我 还是很懵逼。
所以我的诉求是：能够让下面的代码跑起来，给他该给的参数，输出可以实现的结果。关于酬金可以根据任务量再调整。
java和python的环境都已经ok，python编译器是 anaconda 的python3.8，当然下面代码是 python2.*，需要使用2to3转换一下。
库也可以自行下载：pip install cf-clus
包内共8个文件，含有效代码的是3个文件：
第一个： library.py
from subprocess import check_output, CalledProcessError
from tempfile import NamedTemporaryFile
import os
import ConfigParser
import json
import subprocess
import StringIO
def clus(input_dict):
# First we write the ARFF file and the settings file into temporary files.
temporary_arff = NamedTemporaryFile(suffix='.arff', delete=False)
temporary_arff.write(input_dict['arff'])
temporary_arff.close()
temporary_settings = NamedTemporaryFile(suffix='.s', delete=False)
settings = input_dict['settings']
if settings is None:
settings = ''
temporary_settings.write(settings)
temporary_settings.close()
# We need to change the filenames in the settings.
settings = ConfigParser.RawConfigParser()
settings.optionxform = str
settings.read(temporary_settings.name)
if not settings.has_section('Data'):
settings.add_section('Data')
settings.set('Data', 'File', temporary_arff.name)
has_prune = False
# We check if there is prune set data.
if input_dict.get('prune', None) is not None:
temporary_validation = NamedTemporaryFile(suffix='.arff', delete=False)
temporary_validation.write(input_dict['prune'])
temporary_validation.close()
settings.set('Data', 'PruneSet', temporary_validation.name)
has_prune = True
has_test = False
# We check if there is test set data.
if input_dict.get('test', None) is not None:
temporary_test = NamedTemporaryFile(suffix='.arff', delete=False)
temporary_test.write(input_dict['test'])
temporary_test.close()
settings.set('Data', 'TestSet', temporary_test.name)
has_test = True
# We need to enable ClowdFlows output.
if not settings.has_section('Output'):
settings.add_section('Output')
settings.set('Output', 'OutputClowdFlowsJSON', 'Yes')
temporary_settings = open(temporary_settings.name, mode='wb')
settings.write(temporary_settings)
temporary_settings.close()
# Execute CLUS.
args = input_dict['args'].replace(";", "").replace("|", "")
if len(args.strip()) > 0:
args = args.split(" ")
else:
args = []
clus_path = os.path.join(os.path.abspath(os.path.dirname(__file__)), 'bin', 'Clus.jar')
p = subprocess.Popen(["java", "-jar", clus_path] + args + [temporary_settings.name,], stdout=subprocess.PIPE,
stderr=subprocess.PIPE)
if p.returncode != 0 and p.returncode is not None:
raise Exception("There was an error when running CLUS: " + str(p.stderr.read()) + " (Error code: " + str(
p.returncode) + ")")
output = p.stdout.read()
error = p.stderr.read()
if len(error.strip()) > 0:
if "Error: " in error:
raise Exception(error.strip())
try:
output_file = open(temporary_settings.name.replace(".s", ".out"), 'rb')
output = output_file.read()
os.unlink(output_file.name)
except:
pass
fimp = ""
try:
fimp_file = open(temporary_settings.name.replace(".s", ".fimp"), 'rb')
fimp = fimp_file.read()
os.unlink(fimp_file.name)
except:
pass
try:
json_file = open(temporary_settings.name.replace(".s", ".json"), 'rb')
json_contents = json.loads(json_file.read())
returned_settings = json_contents['settings']
models = json_contents['models']
os.unlink(json_file.name)
default = {}
original = {}
pruned = {}
for m in models:
# print m
if m['name'] == 'Default':
default = m['representation']
if m['name'] == 'Original':
original = m['representation']
if m['name'] == 'Pruned':
pruned = m['representation']
except:
returned_settings = None
models = None
default = None
original = None
pruned = None
# We remove all temporary files.
os.unlink(temporary_arff.name)
os.unlink(temporary_settings.name)
if has_prune:
os.unlink(temporary_validation.name)
if has_test:
os.unlink(temporary_test.name)
return {
'output': output,
'settings': returned_settings,
'models': models,
'default': default,
'original': original,
'pruned': pruned,
'fimp': fimp,
# 'error:': error
}
def clus_display_svg(input_dict):
return {}
def clus_display_tree(input_dict):
return {}
def clus_display_tree_and_examples(input_dict):
if type(input_dict['classifier']) == list:
raise Exception("This widget does not work on multiple trees.")
return {}
def clus_display_tree_and_summary(input_dict):
return {}
def handle_setting(name, input_dict, section, settings, checkbox=False):
if not checkbox and input_dict.get(name, None) is not None \
and input_dict.get(name, "").strip() != "" \
and input_dict.get(name, "") != "null":
if not settings.has_section(section):
settings.add_section(section)
settings.set(section, name, input_dict[name])
if checkbox:
if input_dict.get(name, None) is not None \
and input_dict.get(name, "").strip() != "" \
and input_dict.get(name, "") != "null":
if not settings.has_section(section):
settings.add_section(section)
settings.set(section, name, "Yes")
else:
if not settings.has_section(section):
settings.add_section(section)
settings.set(section, name, "No")
def clus_generate_settings(input_dict):
settings = ConfigParser.RawConfigParser()
settings.optionxform = str
settings_buffer = StringIO.StringIO()
handle_setting("RandomSeed", input_dict, "General", settings)
handle_setting("Target", input_dict, "Attributes", settings)
handle_setting("Clustering", input_dict, "Attributes", settings)
handle_setting("Disable", input_dict, "Attributes", settings)
handle_setting("Key", input_dict, "Attributes", settings)
handle_setting("Weights", input_dict, "Attributes", settings)
handle_setting("MinimalWeight", input_dict, "Model", settings)
handle_setting("FTest", input_dict, "Tree", settings)
handle_setting("SplitSampling", input_dict, "Tree", settings)
handle_setting("Heuristic", input_dict, "Tree", settings)
handle_setting("PruningMethod", input_dict, "Tree", settings)
handle_setting("InductionOrder", input_dict, "Tree", settings)
handle_setting("EntropyType", input_dict, "Tree", settings)
handle_setting("BranchFrequency", input_dict, "Output", settings, checkbox=True)
settings.write(settings_buffer)
return {
'settings': settings_buffer.getvalue()
}
第2个 visualization_views.py
from django.shortcuts import render
import os
from utils import clus_tree_to_dot, clus_tree_to_node_edge, get_instance_nodes
from random import random
from django.conf import settings
import arff
def clus_display_svg(request, input_dict, output_dict, widget):
"""Visualization displaying a decision tree"""
import subprocess
from mothra.settings import MEDIA_ROOT
from workflows.helpers import ensure_dir
img_type = 'svg'
if input_dict['img_type'] == 'raster':
img_type = 'png'
dot_text = """digraph J48Tree {
N0 [label="f8" ]
N0->N1 [label="= +"]
N1 [label="f99" ]
N1->N2 [label="= +"]
N2 [label="east (10.0/1.0)" shape=box style=filled ]
N1->N3 [label="= -"]
N3 [label="west (3.0/1.0)" shape=box style=filled ]
N0->N4 [label="= -"]
N4 [label="west (7.0)" shape=box style=filled ]
}"""
if type(input_dict['classifier']) == list:
dot_text = ""
starting_id = 0
for cls in input_dict['classifier']:
dot_representation, starting_id = clus_tree_to_dot(cls['representation'], starting_id)
dot_text += dot_representation + "\n"
# dot_text = dot_text + "digraph " + cls['name'] + " {\n" + \
# dot_representation + "}\n\n"
dot_text = "digraph Tree {\n" + dot_text + "}"
else:
dot_text = "digraph Tree {\n" + clus_tree_to_dot(input_dict['classifier'], 0)[0] + "}"
filename = '/'.join([str(request.user.id), 'decisionTree-clus-%d.dot' % widget.id])
dotfile = filename
destination_dot = '/'.join([MEDIA_ROOT, filename])
ensure_dir(destination_dot)
with open(destination_dot, 'w') as dot_file:
dot_file.write(dot_text)
# png/svg file
filename = '/'.join([str(request.user.id),
'decisionTree-clus-%d.%s' % (widget.id, img_type)
])
destination_img = os.path.join(MEDIA_ROOT, filename)
ensure_dir(destination_img)
try:
dot_path = settings.DOT_PATH
except:
dot_path = 'dot'
subprocess.call(dot_path + " -T%s %s -o %s" % (img_type, destination_dot, destination_img), shell=True)
return render(request,
'visualizations/cf_clus_display_svg_tree.html',
{'filename': filename,
'dotfile': dotfile,
'random': int(random() * 10000000),
'widget': widget,
'input_dict': input_dict})
def clus_display_tree(request, input_dict, output_dict, widget):
"""Visualization displaying a decision tree"""
if type(input_dict['classifier']) == list:
nodes, edges = [], []
starting_id = 0
for cls in input_dict['classifier']:
new_nodes, new_edges, starting_id = clus_tree_to_node_edge(cls['representation'], starting_id)
nodes += new_nodes
edges += new_edges
else:
nodes, edges, index = clus_tree_to_node_edge(input_dict['classifier'], 0)
return render(request,
'visualizations/cf_clus_display_tree.html',
{
'widget': widget,
'input_dict': input_dict,
'nodes': nodes,
'edges': edges
})
def clus_display_tree_and_examples(request, input_dict, output_dict, widget):
"""Visualization displaying a decision tree and the examples in the tree"""
nodes, edges, index = clus_tree_to_node_edge(input_dict['classifier'], 0)
data = arff.loads(input_dict['arff'])
datanodes = []
for instance in data['data']:
instance_nodes = get_instance_nodes(input_dict['classifier'], instance, data['attributes'])
datanodes.append({'data': instance, 'nodes': instance_nodes})
return render(request,
'visualizations/cf_clus_display_tree_and_examples.html',
{
'widget': widget,
'input_dict': input_dict,
'nodes': nodes,
'edges': edges,
'data': data,
'datanodes': datanodes,
'random': int(random() * 10000000),
})
def clus_display_tree_and_summary(request, input_dict, output_dict, widget):
"""Visualization displaying a decision tree and the summary"""
if type(input_dict['classifier']) == list:
nodes, edges = [], []
starting_id = 0
attributes = input_dict['classifier'][0]['representation']['summary']['names']
for cls in input_dict['classifier']:
new_nodes, new_edges, starting_id = clus_tree_to_node_edge(cls['representation'], starting_id)
nodes += new_nodes
edges += new_edges
else:
nodes, edges, index = clus_tree_to_node_edge(input_dict['classifier'], 0)
attributes = input_dict['classifier']['summary']['names']
return render(request,
'visualizations/cf_clus_display_tree_and_summary.html',
{
'widget': widget,
'input_dict': input_dict,
'nodes': nodes,
'edges': edges,
'attributes': attributes,
'random': int(random() * 10000000),
})
第3 个： utils.py
def clus_tree_to_dot(node, node_index):
text = ""
node['dot_id'] = node_index
if 'children' in node:
text += "N" + str(node_index) + ' [label="' + node['test_string'] + '"]\n'
node_index += 1
for child in node['children']:
child_text, node_index = clus_tree_to_dot(child, node_index)
text += child_text
text += "N" + str(node['dot_id']) + "->N" + str(child['dot_id']) + '[label="' + child[
'branch_label'] + '"]' + "\n"
else:
return "N" + str(node_index) + ' [label="' + node[
'target_stat'].replace(',', ',\\n') + '" shape=box style=filled ]\n', node_index + 1
return text, node_index
def clus_tree_to_node_edge(node, node_index):
nodes = []
edges = []
node['dot_id'] = node_index
if 'children' in node:
nodes.append({'id': node['dot_id'], 'label': node['test_string'], 'shape': 'ellipse',
'target_stat': node['target_stat'].replace(',', ',\\n'),
'title': node['target_stat'].replace(',', ',<br>'),
'min': node['summary']['min'],
'max': node['summary']['max'],
'stddev': node['summary']['stddev'],
'avg': node['summary']['avg']})
node_index += 1
for child in node['children']:
child_nodes, child_edges, node_index = clus_tree_to_node_edge(child, node_index)
nodes += child_nodes
edges += child_edges
edges.append({'from': node['dot_id'], 'to': child['dot_id'], 'label': child['branch_label']})
else:
nodes.append({'id': node_index, 'label': node['target_stat'].replace(',', ',\\n'), 'shape': 'box',
'target_stat': node['target_stat'], 'title': node['target_stat'].replace(',', ',<br>'),
'min': node['summary']['min'],
'max': node['summary']['max'],
'stddev': node['summary']['stddev'],
'avg': node['summary']['avg']})
return nodes, edges, node_index + 1
return nodes, edges, node_index
def perform_test(test_string, instance, attributes):
if '(' in test_string:
test_string = test_string.split(' (')[0]
if ' > ' in test_string:
name_value = test_string.split(' > ')
name = name_value[0]
value = float(name_value[1])
i = 0
for a in attributes:
if a[0] == name:
return instance[i] > value
i += 1
if ' = ' in test_string:
name_value = test_string.split(' = ')
name = name_value[0]
value = name_value[1]
i = 0
for a in attributes:
if a[0] == name:
return instance[i] == value
i += 1
if " <= " in test_string:
name_value = test_string.split(' <= ')
name = name_value[0]
value = name_value[1]
i = 0
for a in attributes:
if a[0] == name:
return instance[i] <= value
i += 1
if " in " in test_string:
name_value = test_string.split(' in ')
name = name_value[0]
value = name_value[1]
value = value.replace("{", "")
value = value.replace("}", "")
values = value.split(",")
values = [x.strip() for x in values]
i = 0
for a in attributes:
if a[0] == name:
return instance[i] in values
i += 1
return None
def get_instance_nodes(node, instance, attributes):
node_ids = []
node_ids.append(node['dot_id'])
if 'children' in node:
test_result = perform_test(node['test_string'], instance, attributes)
for child in node['children']:
if test_result == True and child['branch_label'] == 'Yes':
node_ids = node_ids + get_instance_nodes(child, instance, attributes)
if test_result == False and child['branch_label'] == 'No':
node_ids = node_ids + get_instance_nodes(child, instance, attributes)
return node_ids
第4个： settings.py
import os
# === STANDARD PACKAGE SETTINGS ===
PACKAGE_ROOT = os.path.dirname(__file__)
# === AUTO IMPORT OPTIONS ===
#If auto_import_package_data is true then given data file is automatically imported when ClowdFlows project is newly deployed or refreshed from git
AUTO_IMPORT_DB = False
#For auto_import_package_data_replace_option description see the 'replace' option in workflows/import_package command
AUTO_IMPORT_DB_REPLACE_OPTION = True
#If file(s) other than ./db/package_data.json should be imported, auto_import_package_data_files should be corrected
AUTO_IMPORT_DB_FILES = [os.path.join(PACKAGE_ROOT,'db/package_data.json')]
第5个： urls.py
from django.conf.urls.defaults import patterns, include, url
urlpatterns = patterns('',
#url(r'^get-adc-index/widget(?P<widget_id>[0-9]+)/nx/Index.html$', 'workflows.latino.views.get_adc_index', name='get adc index'),
#url(r'^get-adc-index/widget(?P<widget_id>[0-9]+)/(?P<narrow_doc>n?)x/Index.html$', 'workflows.latino.views.get_adc_index', name='get adc index'),
#url(r'^get-adc-index/widget(?P<widget_id>[0-9]+)/(?P<narrow_doc>n?)x/Index(?P<document_id_from>[0-9]+)-(?P<document_id_to>[0-9]+).html$', 'workflows.latino.views.get_adc_index', name='get adc index'),
#url(r'^get-adc-index/widget(?P<widget_id>[0-9]+)/(?P<narrow_doc>n?)x/Document(?P<document_id>[0-9]+).html', 'workflows.latino.views.get_adc_page', name='get adc page'),
)
第6个：interaction_views.py 就一行
