def test_tree():
text = '''
年龄, 工作, 已婚, 信用, 可以借贷?
青年, no , no , 一般 , 不给
青年, no , no , 好 , 不给
青年, yes, no , 好 , 给
青年, yes, yes, 一般 , 给
青年, no , no , 一般 , 不给
中年, no , no , 一般 , 不给
中年, no , no , 好 , 不给
中年, yes, yes, 好 , 给
中年, no , yes, 非常好 , 给
中年, no , yes, 非常好 , 给
老年, no , yes, 非常好 , 给
老年, no , yes, 好 , 给
老年, yes, no , 好 , 给
老年, yes, no , 非常好 , 给
老年, no , no , 一般 , 不给
'''
data, title = datamatrix(text)
dt = DecisionTree(data, labels=title)
n = dt.best_split_feature(data)
tree = dt.decision_tree
print(n)
puts(tree)
test1 = dt.classify(['青年', 'no', 'no', '好'], decision_tree=tree)
test2 = dt.classify(['青年', 'no', 'yes', '非常好'], decision_tree=tree)
test3 = dt.classify(['老年', 'yes', 'yes', '一般'], decision_tree=tree)
test4 = dt.classify(['老年', 'no', 'no', '好'], decision_tree=tree)
test1 | should.eq('不给')
test2 | should.eq('给')
test3 | should.eq('给')
test4 | should.eq('不给')
def brute_force_params(run_func, logger, **params_range):
'''尝试参数的各种可能组合
每个输入参数是一个可能值的列表, 比如 range(0, 250, 10)
使用 itertools.product 遍历所有可能的参数组'''
from itertools import product
import time
total_start = time.time()
total = []
params_product = list(product(*[[(k, elem) for elem in v] for k, v in params_range.items()]))
puts('params prepared, {} variantions'.format(len(params_product)))
for i, params in enumerate(params_product):
loop_start = time.time()
result = run_func(**dict(params))
# result = 0
total.append([params, result])
loop_cost = time.time() - loop_start
total_cost = time.time() - total_start
percent = (i+1) / len(params_product)
info = 'get <{result:.3%}> by: {params} \n [current {loop_cost:.1f}s / total {total_cost:.1f}s] {percent:.2%}'.format_map(vars())
logger.debug(info)
puts('all done!')
logger.debug(total)
return total
def test_word_filler_render():
t1 = os.getcwd() + '/test/test_templates/test_{{name}}_面积计算表.doc'
t2 = os.getcwd() + '/test/test_templates/test_no_field_面积计算表.doc'
from infotext import InfoText
text = '''
单位名称: 测试单位
name: 测试单位name
项目名称: 测试项目
项目编号: 2015-项目编号-001
面积90: 12345.600
面积80: 12345.300
地籍号: 1234567890010010000
四至: 测试路1;测试街2;测试路3;测试街4
土地坐落: 测试路以东,测试街以南
area: 1000
已设定值: value
'''
info = InfoText.from_string(text)
filler = Filler(template_path=t1, output_folder=os.getcwd() + '/test/test_output')
filler.detect_required_fields(unique=False) | puts()
filler = Filler(template_path=t1, output_folder=os.getcwd() + '/test/test_output')
filler.detect_required_fields(unique=True) | puts()
filler = Filler(template_path=t1, output_folder=os.getcwd() + '/test/test_output')
filler.render(info=info)
def test_tree():
text = '''
年龄, 工作, 已婚, 信用, 可以借贷?
青年, no , no , 一般 , 不给
青年, no , no , 好 , 不给
青年, yes, no , 好 , 给
青年, yes, yes, 一般 , 给
青年, no , no , 一般 , 不给
中年, no , no , 一般 , 不给
中年, no , no , 好 , 不给
中年, yes, yes, 好 , 给
中年, no , yes, 非常好 , 给
中年, no , yes, 非常好 , 给
老年, no , yes, 非常好 , 给
老年, no , yes, 好 , 给
老年, yes, no , 好 , 给
老年, yes, no , 非常好 , 给
老年, no , no , 一般 , 不给
'''
data, title = datamatrix(text)
dt = DecisionTree(data, labels=title)
n = dt.best_split_feature(data)
tree = dt.decision_tree
print(n)
puts(tree)
test1 = dt.classify(['青年', 'no', 'no', '好'], decision_tree=tree)
test2 = dt.classify(['青年', 'no', 'yes', '非常好'], decision_tree=tree)
test3 = dt.classify(['老年', 'yes', 'yes', '一般'], decision_tree=tree)
test4 = dt.classify(['老年', 'no', 'no', '好'], decision_tree=tree)
test1 | should.eq('不给')
test2 | should.eq('给')
test3 | should.eq('给')
test4 | should.eq('不给')
def brute_force_params(run_func, logger, **params_range):
'''尝试参数的各种可能组合
每个输入参数是一个可能值的列表, 比如 range(0, 250, 10)
使用 itertools.product 遍历所有可能的参数组'''
from itertools import product
import time
total_start = time.time()
total = []
params_product = list(
product(*[[(k, elem) for elem in v] for k, v in params_range.items()]))
puts('params prepared, {} variantions'.format(len(params_product)))
for i, params in enumerate(params_product):
loop_start = time.time()
result = run_func(**dict(params))
# result = 0
total.append([params, result])
loop_cost = time.time() - loop_start
total_cost = time.time() - total_start
percent = (i + 1) / len(params_product)
info = 'get <{result:.3%}> by: {params} \n [current {loop_cost:.1f}s / total {total_cost:.1f}s] {percent:.2%}'.format_map(
vars())
logger.debug(info)
puts('all done!')
logger.debug(total)
return total
def test_intepolate_polyline_dx():
'''多段线加密 and 分割为多段线的碎片'''
cad = AutoCAD()
for pl in cad.selecting():
# distance = 0.3
distance = 0.2
interpl, report = cad.interpolate_polyline(pl, distance=distance, delete_original=True, break_at_vertexes=True)
puts('多段线加密 分割为多段线的碎片')
def test_info_nested_by_yaml_load():
path = os.getcwd() + '/test/nested.inf'
info = InfoText.from_yaml(path)
puts(info)
print('----')
puts(info.content)
print('----')
print(yaml.dump(info.content))
def test_trim_result():
path = 'result_by_self_train[28000].csv'
result = []
for i, r, b in load_csv(path, sample=0):
result.append(int(r) if r != 'classify failed' else random.choice(list(range(0, 10))))
puts(statistic(result))
result = [[r] for r in result]
write_csv('submit_' + path, rows=result, headers=['val'])
def test_check_result():
path_result = 'result_by_self_train[28000].csv'
path_benchmark = 'rf_benchmark.csv'
result = load_csv(path_result, sample=0)
# benchmark = load_csv(path_benchmark, sample=21)
score = []
for i, r, b in result:
score.append('+' if str(r) == b[1:2] else '-')
puts(statistic(score))
def test_intepolate_polyline():
'''多段线加密'''
cad = AutoCAD()
for pl in cad.selecting():
# distance = 0.3
distance = 0.2
interpl, report = cad.interpolate_polyline(pl, distance=distance, delete_original=True, break_at_vertexes=False)
interpl.color = 'green'
puts('多段线加密')
def test_check_result():
path_result = 'result_by_self_train[28000].csv'
path_benchmark = 'rf_benchmark.csv'
result = load_csv(path_result, sample=0)
# benchmark = load_csv(path_benchmark, sample=21)
score = []
for i, r, b in result:
score.append('+' if str(r) == b[1:2] else '-')
puts(statistic(score))
def test_trim_result():
path = 'result_by_self_train[28000].csv'
result = []
for i, r, b in load_csv(path, sample=0):
result.append(
int(r) if r != 'classify failed' else random.
choice(list(range(0, 10))))
puts(statistic(result))
result = [[r] for r in result]
write_csv('submit_' + path, rows=result, headers=['val'])
def test_info_additional_keys():
from pyshould import should
path = os.getcwd() + '/test/nested.inf'
info = InfoText.from_yaml(path)
puts(info.content)
info.get('a') | should.eq(123)
info.get('ErrorKey') | should.eq(None)
info.get('foo') | should.eq('bar') # from key<default>
info.get('current_year') | should.eq(1404) # key<default> contains this 1404
print(info.get('current_date')) # key<default> does not contain this,
def test_redraw_vertex_sequence():
'''重绘多段线的顶点顺序'''
cad = AutoCAD()
for pl in cad.selecting():
first = None # first = -1
redrawpl = cad.redraw_vertex_sequence(pl, first=first,
hint=False,
auto_reverse=True)
redrawpl.color = 'yellow'
puts('重绘顶点顺序')
def test_de_intepolate_polyline():
'''多段线抽稀'''
cad = AutoCAD()
for pl in cad.selecting():
interpl, report = cad.de_interpolate_polyline(pl, threshold=3)
if abs(interpl.area - pl.area) < 30:
interpl.color = 'green'
pl.delete()
else:
interpl.delete()
puts('多段线抽稀 report=')
def parse_packaging_served_detail_page(url, folder):
url | puts()
r = requests.get(url)
doc = PyQuery(r.text) | puts()
# urls = doc('#project-modules li.image>img').attr('src') | puts()
folder = to_file_name(folder) | puts()
if not os.path.exists(folder):
os.mkdir(folder)
for i, elem in enumerate(doc("#project-modules li.image>img"), 1):
url = PyQuery(elem).attr("src")
save_image(url, folder, folder + "_" + str(i))
def train(self):
for i in range(self.n_trees):
features_index = random.sample(list(range(self.n_features)), self.n_features_per_tree)
self.features_index_for_trees.append(features_index)
data_filtered = self.filter_traindata(features_index)
labels_filtered = self.filter_labels(features_index)
tree = DecisionTree(data_filtered, labels=labels_filtered)
self.trees.append(tree)
if self.verbose:
puts('train done: total {self.n_rows} rows, {self.n_trees} trees, each tree use {self.n_rows_per_tree} rows data')
t1, t2, t3, *_, tn = self.trees
pprint(t1.decision_tree)
pprint(t2.decision_tree)
pprint(t3.decision_tree)
pprint('------------------------')
def test_documasonry_generate():
template_paths = [os.getcwd() + '/test/test_templates/_test_{{项目名称}}-申请表.xls',
os.getcwd() + '/test/test_templates/test_{{name}}_面积计算表.doc',
os.getcwd() + '/test/test_templates/test_{{测试单位}}-宗地图.dwg',
os.getcwd() + '/test/test_templates/test_no_field_面积计算表.doc',
]
output_path = os.getcwd() + '/test/test_output'
masonry = Documasonry(output_path=output_path, template_paths=template_paths)
text = '''
项目名称: test1
单位名称: test2
地籍号: 110123122
name: sjgisdgd
面积90: 124.1
面积80: 234.2
area: 124.2
测试单位: testconm
title: testtitle
project: pro.
date: 20124002
ratio: 2000
landcode: 235
area80: 94923
area90: 3257
'''
info = InfoText.from_string(text)
masonry.generate(info=info, save=True, add_index=True) | puts()
def points_list_to_variant(self, coord):
import pylon
data = list(float(x) for x in pylon.flatten(coord)) | pylon.puts()
if len(data) % 2 == 1:
raise AttributeError("point_to_variant: coord length must be even")
com_seq = win32com.client.VARIANT(pythoncom.VT_ARRAY | pythoncom.VT_R8, data)
return com_seq
def save_image(url, folder, image_name=None):
if not image_name:
file_name = folder + "/" + url.split("/")[-1]
else:
ext = url.split(".")[-1]
file_name = folder + "/" + to_file_name(image_name) + "." + ext
if os.path.exists(file_name):
puts("-- already exist file_name")
return
r = requests.get(url, stream=True)
if r.status_code == 200:
with open(file_name, "wb") as f:
r.raw.decode_content = True
shutil.copyfileobj(r.raw, f)
puts("save done! url >> file_name")
def test_documasonry_detect_fields():
template_paths = [os.getcwd() + '/test/test_templates/_test_{{项目名称}}-申请表.xls',
os.getcwd() + '/test/test_templates/test_{{name}}_面积计算表.doc',
os.getcwd() + '/test/test_templates/test_{{测试单位}}-宗地图.dwg',
os.getcwd() + '/test/test_templates/test_no_field_面积计算表.doc',
]
output_path = os.getcwd() + '/test/test_output'
masonry = Documasonry(output_path=output_path, template_paths=template_paths)
masonry.detect_required_fields() | puts()
def save(self, info, close=True, prefix=''):
self.output_name = prefix + evalute_field(os.path.basename(self.template_path), info)
output_path = os.path.join(self.output_folder, self.output_name)
output_path = output_path.replace('\\', '/')
output_path = output_path.replace('/', '\\')
if os.path.exists(output_path):
fix = time.strftime('.backup-%Y%m%d-%H%M%S')
os.rename(output_path, fix.join(os.path.splitext(output_path)))
try:
self.document.SaveAs(output_path)
puts('save document done - output_path')
except Exception:
raise
t = 'Word Filler can not save document: <{}>'.format(output_path)
raise SaveDocumentError(t)
if close:
self.document.Close()
def test_scan_entities():
'''扫描选中对象的信息
对于多段线额外统计 闭合线总面积 开放线总长度
hole: 将多段线视为外部包裹线和内部孔洞, 统计面积时以最大面积减去其他较小的
'''
r = cad.scan_entities(selecting_entities,
hole=True, error_color=None, error_layer=None)
for line in r:
line | puts()
def train(self):
for i in range(self.n_trees):
features_index = random.sample(list(range(self.n_features)),
self.n_features_per_tree)
self.features_index_for_trees.append(features_index)
data_filtered = self.filter_traindata(features_index)
labels_filtered = self.filter_labels(features_index)
tree = DecisionTree(data_filtered, labels=labels_filtered)
self.trees.append(tree)
if self.verbose:
puts(
'train done: total {self.n_rows} rows, {self.n_trees} trees, each tree use {self.n_rows_per_tree} rows data'
)
t1, t2, t3, *_, tn = self.trees
pprint(t1.decision_tree)
pprint(t2.decision_tree)
pprint(t3.decision_tree)
pprint('------------------------')
def test_rebuild_arc_polyline():
'''将加密后的poly转为圆弧poly
处理选中的多段线
如果有选中的点, 将这些点作为分隔符'''
cad = AutoCAD()
polylines = []
points = []
for en in cad.selecting():
if en.entity_type == 'Point':
points.append(en)
elif en.entity_type == 'Polyline' and en.closed:
polylines.append(en)
dist = 3
for pl in polylines:
arcpl = cad.rebuild_arc_polyline(pl, threshold=dist,
segment_points=[(p.x, p.y) for p in points])
arcpl.color = 'yellow'
report = '转为圆弧poly 原面积={:.4f} 新面积={:.4f} 相差={:.4f} ({:.4%})'
puts(report.format(pl.area, arcpl.area, arcpl.area - pl.area, (arcpl.area-pl.area)/pl.area))
def parse_incredibal(path):
text = open(path).read()
doc = PyQuery(text)
for elem in doc('article'):
elem = PyQuery(elem)
# elem('div h3>a').text() | puts()
url = elem('figure div img').attr('src') | puts()
# file_name = elem('figure div img').attr('src').split('/')[-1] | puts()
save_image(url, folder='food')
def test_word_filler_detect_fields():
t1 = os.getcwd() + '/test/test_templates/test_{{name}}_面积计算表.doc'
t2 = os.getcwd() + '/test/test_templates/test_no_field_面积计算表.doc'
from infotext import InfoText
yaml_info = InfoText.from_yaml(os.getcwd() + '/test/测试单位.inf')
filler = Filler(template_path=t1, output_folder=os.getcwd() + '/test/test_output')
filler.detect_required_fields(unique=False) | puts()
filler = Filler(template_path=t1, output_folder=os.getcwd() + '/test/test_output')
filler.detect_required_fields(unique=True) | puts()
filler = Filler(template_path=t2, output_folder=os.getcwd() + '/test/test_output')
filler.detect_required_fields(unique=False) | puts()
filler = Filler(template_path=t2, output_folder=os.getcwd() + '/test/test_output')
filler.detect_required_fields(unique=True) | puts()
filler = Filler(template_path=t1, output_folder=os.getcwd() + '/test/test_output')
filler.render(info=yaml_info)
def kmeans(dataSet, k):
numSamples = dataSet.shape[0]
# first column stores which cluster this sample belongs to,
# second column stores the error between this sample and its centroid
clusterAssment = mat(zeros((numSamples, 2)))
clusterChanged = True
## step 1: init centroids
centroids = initCentroids(dataSet, k)
while clusterChanged:
clusterChanged = False
## for each sample
for i in range(numSamples):
minDist = 100000.0
minIndex = 0
## for each centroid
## step 2: find the centroid who is closest
for j in range(k):
distance = euclDistance(centroids[j, :], dataSet[i, :])
if distance < minDist:
minDist = distance
minIndex = j
## step 3: update its cluster
if clusterAssment[i, 0] != minIndex:
clusterChanged = True
clusterAssment[i, :] = minIndex, minDist**2
## step 4: update centroids
for j in range(k):
pointsInCluster = dataSet[nonzero(clusterAssment[:, 0].A == j)[0]]
centroids[j, :] = mean(pointsInCluster, axis = 0)
pylon.puts('Congratulations, cluster complete!')
return centroids, clusterAssment
def test_find_nearest_text():
cad = AutoCAD()
entities = list(cad.selecting())
numbers = []
names = []
for text in entities:
if text.color == 'black':
names.append(text)
else:
numbers.append(text)
from Converter import SpaceCoordinate
dist = SpaceCoordinate().distance2
for number in numbers:
near = min(names, key=lambda name: dist(name.mid_point, number.mid_point)) | puts()
near.color = 'cyan'
def parse_packaging_served(path):
text = open(path).read()
doc = PyQuery(text)
result = []
for elem in doc("li>div"):
elem = PyQuery(elem)
# elem('div h3>a').text() | puts()
# name = elem('.cover-name>a').text() | puts()
title = elem(".cover-img img.cover-img-standard").attr("title")
image = elem(".cover-img img.cover-img-standard").attr("src")
detail = elem(".cover-name a.cover-name-link").attr("href")
result.append((title, image, detail))
# save_image(image, folder='packaging_served_images_preview',
# image_name=title)
# result | puts()
for i, (title, image, detail) in enumerate(result):
if i < 16:
continue
puts("@ start parse folder i title detail")
site = "http://www.packagingserved.com"
parse_packaging_served_detail_page(url=site + detail, folder=title)
def test_jinja_edge_cases():
from infotext import InfoText
text = '''
codes: 1231234000050280000
borders: 空地;空地;空地;空地
'''
template = '''
{{(codes | string)[12:15]}}-01
{{(codes | string)[12:15]}}-01
'''
info = InfoText.from_string(text)
t = Template(template)
result = t.render(**info.content)
result | puts()
def showCluster(dataSet, k, centroids, clusterAssment):
numSamples, dim = dataSet.shape
if dim != 2:
pylon.puts("Sorry! I can not draw because the dimension of your data is not 2!")
return 1
mark = ['or', 'ob', 'og', 'ok', '^r', '+r', 'sr', 'dr', '<r', 'pr']
if k > len(mark):
pylon.puts("Sorry! Your k is too large! please contact Zouxy")
return 1
# draw all samples
for i in range(numSamples):
markIndex = int(clusterAssment[i, 0])
plt.plot(dataSet[i, 0], dataSet[i, 1], mark[markIndex])
mark = ['Dr', 'Db', 'Dg', 'Dk', '^b', '+b', 'sb', 'db', '<b', 'pb']
# draw the centroids
for i in range(k):
plt.plot(centroids[i, 0], centroids[i, 1], mark[i], markersize = 12)
plt.show()
def render(self, info):
self.info = info
self.app.Visible = True
if self.info.get('target_position'):
# 如含有 target_position 字段 编辑前需要调整模板全体 object 位置
target_position = self.info.content['target_position']
if isinstance(target_position, list) and len(target_position) == 4:
# target_center and target_size [566371.2180, 4340932.6223, 202.3, 202.3]
self.fix_position(target_center=target_position[:2],
target_size=target_position[2:])
elif isinstance(target_position, str):
if not os.path.isfile(target_position):
target_position = os.path.join(self.output_folder, target_position)
self.insert_block(dwg_path=target_position)
last_entity = list(self.entities(kinds='BlockReference'))[0] | puts()
target_center = self.mid_point(last_entity)
target_size = self.bounding_box_size(last_entity)
# print(last_entity.)
last_entity.Delete()
self.fix_position(target_center=target_center, target_size=target_size)
for en in self.text_entities():
val = evalute_field(field=en.TextString, info=info)
if val in (None, ''):
raise InfoKeyError('无法找到字段的值 {}'.format(en.TextString))
if en.TextString.startswith('{{') and en.TextString.endswith('dwg}}'):
# block field syntax should insert dwg block
if not os.path.isfile(val):
val = os.path.join(self.output_folder, val)
self.insert_block(dwg_path=val)
en.Delete()
else:
en.TextString = val
self.document.SendCommand('zoom e ')
def test_cad_filler_detect_fields():
t1 = os.getcwd() + '/test/test_templates/test_{{测试单位}}-宗地图.dwg'
filler = Filler(template_path=t1, output_folder=os.getcwd() + '/test/test_output')
filler.detect_required_fields(unique=False) | puts()
filler = Filler(template_path=t1, output_folder=os.getcwd() + '/test/test_output')
filler.detect_required_fields(unique=True) | puts()
def test_remove_same_point_polyline():
'移除poly中的重复节点'
cad = AutoCAD()
for pl in cad.selecting():
plnew, report = cad.remove_same_points_polyline(pl, threshold=0.0001)
puts(report)
