def check_font():
flist = font_manager.findSystemFonts()
names = [
font_manager.FontProperties(fname=fname).get_name() for fname in flist
]
if not (FONT_NAME in names):
font_manager._rebuild()
def getFontStyleList(self, ):
from matplotlib import font_manager
font_manager._rebuild()
# 리스트의 원소(폰트파일의 경로)만큼 반복
fontDict = dict({})
for v in font_manager.findSystemFonts(fontpaths=None, fontext='ttf'):
try:
# 각 폰트파일의 경로를 사용하여 폰트 속성 객체 얻기
fprop = font_manager.FontProperties(fname=v)
# 폰트 속성중 이름과 파일 경로를 딕셔러리로 구성하여 리스트에 추가.
fontDict[fprop.get_name()] = fprop.get_file()
except:
continue
fontDict = dict(sorted(fontDict.items())) #sort key
newFontDict = dict({})
for key in fontDict.keys():
newFontDict["""<span style="font-family:""" + key + """;">""" +
key + "</span>"] = fontDict[
key] #apply font style on display
fontDict = newFontDict
#<span style="font-family:Ami R;">Ami R</span>
fontDict["NotoSans"] = "./font/NotoSansKR-Regular.otf"
return newFontDict
def chinese_setting(
url="http://cnbj1.fds.api.xiaomi.com/browser-algo-nanjing/data/SimHei.ttf"
):
"""
:param url: SimHei字体下载链接
:return:
"""
print('开始设置中文...')
matplotlibrc_path = Path(matplotlib.matplotlib_fname())
ttf_path = matplotlibrc_path.parent.__str__() + '/fonts/ttf'
ttf_url = 'https://raw.githubusercontent.com/Jie-Yuan/Jie-Yuan.github.io/master/SimHei.ttf' if url is None else url
if list(Path(ttf_path).glob('SimHei.ttf')):
pass
else:
print('下载字体...')
os.popen("cd %s && wget %s" % (ttf_path, ttf_url)).read()
os.popen("rm -rf ~/.matplotlib/*.cache").read()
print('设置字体...')
setting1 = 'font.family: sans-serif'
setting2 = 'font.sans-serif: SimHei, Bitstream Vera Sans, Lucida Grande, Verdana, Geneva, Lucid, Arial, Helvetica, Avant Garde, sans-serif'
setting3 = 'axes.unicode_minus: False'
os.system('echo > %s' % matplotlibrc_path)
os.system('echo %s >> %s' % (setting1, matplotlibrc_path))
os.system('echo %s >> %s' % (setting2, matplotlibrc_path))
os.system('echo %s >> %s' % (setting3, matplotlibrc_path))
_rebuild()
print('请重启kernel测试...')
def test_font_priority():
with rc_context(rc={"font.sans-serif": ["cmmi10", "Bitstream Vera Sans"]}):
# force the font manager to rebuild it self
_rebuild()
font = findfont(FontProperties(family=["sans-serif"]))
assert_equal(os.path.basename(font), "cmmi10.ttf")
# force it again
_rebuild()
def colab_kvis():
subprocess.run(["apt-get", "install", "fonts-nanum*"])
path = '/usr/share/fonts/truetype/nanum/NanumBarunGothic.ttf'
font = fm.FontProperties(fname=path, size=12)
plt.rc('font', family=font.get_name())
fm._rebuild()
matplotlib.rcParams['axes.unicode_minus'] = False
return print("한글 시각화 준비 완료")
def initialize_mpl_fonts(self):
from matplotlib.font_manager import _rebuild; _rebuild()
fp = FontProperties(fname=self.font_path)
rcParams['font.serif'] = fp.get_name()
rcParams['font.family'] = 'serif'
def decode(args: Dict[str, str]):
""" Performs decoding on a test set, and save the best-scoring decoding results.
If the target gold-standard sentences are given, the function also computes
corpus-level BLEU score.
@param args (Dict): args from cmd line
"""
print("load test source sentences from [{}]".format(
args['TEST_SOURCE_FILE']),
file=sys.stderr)
test_data_src = read_corpus(args['TEST_SOURCE_FILE'], source='src')
if args['TEST_TARGET_FILE']:
print("load test target sentences from [{}]".format(
args['TEST_TARGET_FILE']),
file=sys.stderr)
test_data_tgt = read_corpus(args['TEST_TARGET_FILE'], source='tgt')
print("load model from {}".format(args['MODEL_PATH']), file=sys.stderr)
model = NMT.load(args['MODEL_PATH'])
if args['--cuda']:
model = model.to(torch.device("cuda"))
hypotheses = beam_search(model,
test_data_src,
beam_size=int(args['--beam-size']),
max_decoding_time_step=int(
args['--max-decoding-time-step']))
if args['TEST_TARGET_FILE']:
top_hypotheses = [hyps[0] for hyps in hypotheses]
bleu_score = compute_corpus_level_bleu_score(test_data_tgt,
top_hypotheses)
print('Corpus BLEU: {}'.format(bleu_score * 100), file=sys.stderr)
with open(args['OUTPUT_FILE'], 'w') as f:
for src_sent, hyps in zip(test_data_src, hypotheses):
top_hyp = hyps[0]
hyp_sent = ' '.join(top_hyp.value)
f.write(hyp_sent + '\n')
if args['--plot-attention']:
plt.rcParams['font.family'] = ['sans-serif']
plt.rcParams['font.sans-serif'] = [
'Arial Unicode MS', 'Arial', 'sans-serif'
]
from matplotlib.font_manager import _rebuild
_rebuild()
output_dir = os.path.dirname(args['OUTPUT_FILE'])
for idx, (src_sent,
hyps) in tqdm(enumerate(zip(test_data_src, hypotheses)),
desc='Plot attention',
file=sys.stdout):
top_hyp = hyps[0]
hyp_sent = top_hyp.value
hyp_att = top_hyp.attention
filename = output_dir + '/att_%d.jpg' % idx
plot_attention(hyp_att, src_sent, hyp_sent, filename)
def draw_flight_graph(weighted_edges_list, node_num, weighted_nodes_list=None):
G = nx.Graph()
G.add_weighted_edges_from(weighted_edges_list, alpha=0.5, edge_color='g')
mfm._rebuild()
mpb.rcParams['font.family'] = ['Heiti TC']
fig = plt.figure(figsize=(8, 8))
ax = fig.add_subplot(121)
pos = nx.spring_layout(G, k=0.25)
node_list, node_sizes = [], []
if weighted_nodes_list != None:
for (node, weight) in weighted_nodes_list:
node_list.append(node)
node_sizes.append(weight)
print(len(node_list), len(node_sizes))
#color codes: #FFD94A (bright yellow); #5BC2E7 (water blue), #4364A2 (darker blue), #D59439 (dark yellow)
##E3C25B (dirt yellow)
nx.draw_networkx_nodes(G,
pos,
nodelist=node_list,
node_size=node_sizes,
node_color='#D59439',
alpha=1)
nx.draw_networkx_nodes(G,
pos,
nodelist=['中国'],
node_size=500,
node_color='#8C1515',
alpha=1)
else:
nx.draw_networkx_nodes(G,
pos,
node_size=500,
node_color=range(1, node_num + 1),
alpha=1,
cmap=plt.cm.Blues)
nx.draw_networkx_nodes(G,
pos,
nodelist=['中国'],
node_size=300,
node_color='r',
alpha=1)
for edge in weighted_edges_list:
nx.draw_networkx_edges(G,
pos,
edgelist=[edge],
width=edge[2] * 100,
alpha=1,
edge_color='#4364A2')
nx.draw_networkx_labels(G,
pos,
font_size=7,
font_family='Heiti TC',
font_weight='bold')
plt.axis('off')
plt.show()
def resolve_chinese_problem():
# 中文乱码处理
# https://www.zhihu.com/question/25404709
from matplotlib.font_manager import _rebuild
_rebuild() # reload一下
# 指定默认字体
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['font.family'] = 'sans-serif'
# 用来正常显示负号
plt.rcParams['axes.unicode_minus'] = False
def plot_bar(prop, cat_var, freq):
mfm._rebuild()
# n = len(mfm.fontManager.ttflist)
mpb.rcParams['font.family'] = ['Heiti TC']
width = 0.35
fig = plt.figure()
ax = fig.add_subplot(111)
rect = ax.bar(cat_var, freq, width, color='#4364A2', label=freq)
ax.set_ylabel('数量')
ax.set_xlabel(prop)
ax.set_title(prop + "数量柱状图")
autolabel(ax, rect)
plt.show()
def getFontStyleList(self, ):
font_manager._rebuild()
# 리스트의 원소(폰트파일의 경로)만큼 반복
fontDict = dict()
for v in font_manager.findSystemFonts(fontpaths=None, fontext='ttf'):
try:
# 각 폰트파일의 경로를 사용하여 폰트 속성 객체 얻기
fprop = font_manager.FontProperties(fname=v)
# 폰트 속성중 이름과 파일 경로를 딕셔러리로 구성하여 리스트에 추가.
fontDict[fprop.get_name()] = fprop.get_file()
except:
continue
fontDict["NotoSans"] = "./font/NotoSansKR-Regular.otf"
return fontDict
def test_find_ttc():
fp = FontProperties(family=["WenQuanYi Zen Hei"])
if Path(findfont(fp)).name != "wqy-zenhei.ttc":
if not os.environ.get("TRAVIS") or sys.platform != "linux":
pytest.skip("Font may be missing")
# Travis appears to fail to pick up the ttc file sometimes. Try to
# rebuild the cache and try again.
fm._rebuild()
assert Path(findfont(fp)).name == "wqy-zenhei.ttc"
fig, ax = plt.subplots()
ax.text(.5, .5, "\N{KANGXI RADICAL DRAGON}", fontproperties=fp)
for fmt in ["raw", "svg", "pdf", "ps"]:
fig.savefig(BytesIO(), format=fmt)
def test_find_ttc():
fp = FontProperties(family=["WenQuanYi Zen Hei"])
if Path(findfont(fp)).name != "wqy-zenhei.ttc":
# Travis appears to fail to pick up the ttc file sometimes. Try to
# rebuild the cache and try again.
fm._rebuild()
assert Path(findfont(fp)).name == "wqy-zenhei.ttc"
fig, ax = plt.subplots()
ax.text(.5, .5, "\N{KANGXI RADICAL DRAGON}", fontproperties=fp)
fig.savefig(BytesIO(), format="raw")
fig.savefig(BytesIO(), format="svg")
with pytest.raises(RuntimeError):
fig.savefig(BytesIO(), format="pdf")
with pytest.raises(RuntimeError):
fig.savefig(BytesIO(), format="ps")
def test_find_ttc():
fp = FontProperties(family=["WenQuanYi Zen Hei"])
if Path(findfont(fp)).name != "wqy-zenhei.ttc":
# Travis appears to fail to pick up the ttc file sometimes. Try to
# rebuild the cache and try again.
fm._rebuild()
assert Path(findfont(fp)).name == "wqy-zenhei.ttc"
fig, ax = plt.subplots()
ax.text(.5, .5, "\N{KANGXI RADICAL DRAGON}", fontproperties=fp)
fig.savefig(BytesIO(), format="raw")
fig.savefig(BytesIO(), format="svg")
with pytest.raises(RuntimeError):
fig.savefig(BytesIO(), format="pdf")
with pytest.raises(RuntimeError):
fig.savefig(BytesIO(), format="ps")
def plot_with_labels(low_dim_embs, labels, filename):
assert low_dim_embs.shape[0] >= len(
labels), 'More labels than embeddings'
_rebuild() # reload一下
plt.rcParams['font.sans-serif'] = ['SimHei'] # 用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False # 用来正常显示负号
plt.figure(figsize=(18, 18)) # in inches
for i, label in enumerate(labels):
x, y = low_dim_embs[i, :]
plt.scatter(x, y)
plt.annotate(label,
xy=(x, y),
xytext=(5, 2),
textcoords='offset points',
ha='right',
va='bottom')
plt.savefig(filename)
def setup():
''' Set up. '''
plt.close('all')
original_units_registry = matplotlib.units.registry.copy()
original_settings = matplotlib.rcParams.copy()
ver = easypyplot.util.matplotlib_version_tuple()
# Setup routine introduced from 1.5.
if ver >= (1, 5):
matplotlib.testing.setup()
# Style name has changed over matplotlib versions.
# See changes from <matplotlib repo>/lib/matplotlib/testing/conftest.py:mpl_test_settings()
if ver >= (3, 2):
matplotlib.style.use(['classic', '_classic_test_patch'])
elif ver >= (2, 0):
matplotlib.style.use('_classic_test')
elif ver >= (1, 5):
matplotlib.style.use('classic')
# Times bold bug.
# https://stackoverflow.com/questions/33955900
# https://github.com/matplotlib/matplotlib/issues/5574
#
# The proposed solution simply removes `'roman'` from `weight_dict`, which
# may cause `KeyError` when `weight_dict['roman']` is used in other places.
#
# Instead we order `'roman'` after `'bold'`, so when matching weights,
# "Times New Roman Bold" will first match `'bold'` and get its weight.
#
# Also we need to update the shortcut module method `findfont`. It may
# already be imported in other modules, so instead of update its value, we
# also need to update the font lists of the previous font manager, so that
# those already imported can also use the new font lists.
if ver >= (2, 0):
# Order `'bold'` before `'roman'` in the weight list.
roman_weight = mlpfm.weight_dict.pop('roman', None)
mlpfm.weight_dict = OrderedDict(mlpfm.weight_dict)
if roman_weight is not None:
mlpfm.weight_dict['roman'] = roman_weight
# Rebuild font manager and update existing font lists.
fm = mlpfm.fontManager
mlpfm._rebuild() # pylint: disable=protected-access
mlpfm.findfont = mlpfm.fontManager.findfont
fm.ttflist = mlpfm.fontManager.ttflist
fm.afmlist = mlpfm.fontManager.afmlist
return original_units_registry, original_settings
def visualize2(directory, n):
tag_list = []
track_list = []
for category in ['genre', 'instrument', 'mood_theme']:
tsv_file = os.path.join(directory, category + '.tsv')
data = pd.read_csv(tsv_file, delimiter='\t')
data = data.sort_values(by=['tracks'], ascending=False)
data = data[:n]
tag_list += list(data['tag'])
track_list += list(data['tracks'])
fig = plt.figure(figsize=(12, 2.5))
plt.style.use('seaborn-whitegrid')
font_manager._rebuild()
rcParams['font.family'] = 'serif'
rcParams['font.serif'] = 'Times'
plt.grid(False)
plt.ylabel('# of tracks')
plt.xlim([-1, 60])
plt.ylim([0, 18000])
for i, color in enumerate(['c', 'm', 'y']):
indices = np.arange(n * i, n * (i + 1))
plt.bar(indices, np.array(track_list)[indices], align='center')
for i in [0, 20, 40, 59]:
plt.text(i,
track_list[i] + 300,
track_list[i],
fontsize=8,
horizontalalignment='center')
plt.xticks(np.arange(len(tag_list)), tag_list, rotation='vertical')
ylabels = np.arange(4, 20, 4)
plt.yticks(ylabels * 1000, ['{}k'.format(ylabel) for ylabel in ylabels])
plt.subplots_adjust(bottom=0.4)
output_file = os.path.join(directory, 'top{}.pdf'.format(n))
plt.savefig(output_file, bbox_inches='tight')
# plt.show()
plt.close()
def set_KRFONT(KRFONT: str):
system = platform.system()
if system == 'Windows' or 'Linux':
if system == 'Windows':
datapath = os.getcwd() + '\\'
imagepath = datapath + 'images\\'
# ttf 폰트 전체개수
font_list[:10]
f = [f.name for f in fm.fontManager.ttflist]
f[:10]
[(f.name, f.fname) for f in fm.fontManager.ttflist if 'Nanum' in f.name]
path = 'C:\\Windows\\Fonts\\NanumBarunGothic.ttf'
elif system == 'Linux':
datapath = os.getcwd() + '//'
imagepath = datapath + 'images//'
# !apt-get update -qq
# !apt-get install fonts-nanum* -qq
path = '/usr/share/fonts/truetype/nanum/NanumGothic.ttf'
KRFONT = fm.FontProperties(fname=path, size=10).get_name()
fm._rebuild()
mpl.rcParams['axes.unicode_minus'] = False
KRFONT = fm.FontProperties(fname=path, size=10).get_name()
add_KRFONT(KRFONT)
print(f"# matplotlib 한글 사용 가능: {KRFONT}")
else:
sys.exit('ERROR: Sorry, my code has compatibility with Windows and Linux only.')
return KRFONT
def get_font_family():
"""
시스템 환경에 따른 기본 폰트명을 반환하는 함수
"""
import platform
system_name = platform.system()
# colab 사용자는 system_name이 'Linux'로 확인
if system_name == "Darwin":
font_family = "AppleGothic"
elif system_name == "Windows":
font_family = "Malgun Gothic"
else:
# Linux
# colab에서는 runtime을 <꼭> 재시작 해야합니다.
# 런타임을 재시작 하지 않고 폰트 설치를 하면 기본 설정 폰트가 로드되어 한글이 깨집니다.
import matplotlib.font_manager as fm
fontpath = '/usr/share/fonts/truetype/nanum/NanumBarunGothic.ttf'
font = fm.FontProperties(fname=fontpath, size=9)
fm._rebuild()
font_family = "NanumBarunGothic"
return font_family
def test():
_rebuild()
def printKickDistribution(sp: la_class.ServerParam, feat: la_class.Feature,
path: str) -> None:
import matplotlib
#matplotlib.use('Agg')
from mpl_toolkits.mplot3d import Axes3D
from matplotlib.patches import Circle
import mpl_toolkits.mplot3d.art3d as art3d
from matplotlib import pyplot as plt
import matplotlib.font_manager as fm
xlim: cython.double = sp.pitch_length / 2 + 5.0
ylim: cython.double = sp.pitch_width / 2 + 5.0
fm._rebuild()
#plt.rc('font', family='Times New Roman')
#plt.rcParams['font.family'] = 'Times New Roman'
fig = plt.figure(figsize=(10.5, 6.8))
plt.rcParams['font.family'] = 'Times New Roman'
plt.rcParams['ps.useafm'] = True
plt.rcParams['pdf.use14corefonts'] = True
plt.rcParams['text.usetex'] = True
ax = fig.add_subplot(111, projection='3d', azim=240)
ax.set_xlabel("{\it x}", fontsize=32, labelpad=20)
ax.set_ylabel("{\it y}", fontsize=32, labelpad=20)
ax.set_zlabel("{\it distance}", fontsize=32, labelpad=20)
ax.set_xlim(-xlim, xlim)
ax.set_ylim(ylim, -ylim)
ax.tick_params(labelsize=32, pad=10)
# plot soccer fields
ax.plot3D([sp.goal_line_l, -sp.penalty_area_x],
[-sp.penalty_area_y, -sp.penalty_area_y], [0, 0],
color='g',
linewidth=4)
ax.plot3D([sp.goal_line_l, -sp.penalty_area_x],
[sp.penalty_area_y, sp.penalty_area_y], [0, 0],
color='g',
linewidth=4)
ax.plot3D([sp.goal_line_r, sp.penalty_area_x],
[-sp.penalty_area_y, -sp.penalty_area_y], [0, 0],
color='g',
linewidth=4)
ax.plot3D([sp.goal_line_r, sp.penalty_area_x],
[sp.penalty_area_y, sp.penalty_area_y], [0, 0],
color='g',
linewidth=4)
ax.plot3D([-sp.penalty_area_x, -sp.penalty_area_x],
[sp.penalty_area_y, -sp.penalty_area_y], [0, 0],
color='g',
linewidth=4)
ax.plot3D([sp.penalty_area_x, sp.penalty_area_x],
[sp.penalty_area_y, -sp.penalty_area_y], [0, 0],
color='g',
linewidth=4)
ax.plot3D([sp.goal_line_l, sp.goal_line_r],
[-sp.pitch_width / 2, -sp.pitch_width / 2], [0, 0],
color='g',
linewidth=4)
ax.plot3D([sp.goal_line_l, sp.goal_line_r],
[sp.pitch_width / 2, sp.pitch_width / 2], [0, 0],
color='g',
linewidth=4)
ax.plot3D([sp.goal_line_l, sp.goal_line_l],
[-sp.pitch_width / 2, sp.pitch_width / 2], [0, 0],
color='g',
linewidth=4)
ax.plot3D([sp.goal_line_r, sp.goal_line_r],
[-sp.pitch_width / 2, sp.pitch_width / 2], [0, 0],
color='g',
linewidth=4)
plt.plot([0, 0], [-sp.pitch_width / 2, sp.pitch_width / 2],
color='g',
linewidth=4)
p = Circle((0.5, 0.5), 9, ec="g", fc="None", linewidth=4)
ax.add_patch(p)
art3d.pathpatch_2d_to_3d(p, z=0.0, zdir="z")
# plot kick distribution
for i in range(len(feat.kick_sequence) - 1):
# if not tarminal condition
if ( feat.kick_sequence[i][4] != -1 \
and feat.kick_sequence[i+1][4] != 0 ):
ax.plot3D([feat.kick_sequence[i][1], feat.kick_sequence[i][1]],
[feat.kick_sequence[i][2], feat.kick_sequence[i][2]], [
0,
lib.calcDistC(feat.kick_sequence[i][1],
feat.kick_sequence[i][2],
feat.kick_sequence[i + 1][1],
feat.kick_sequence[i + 1][2])
], "red")
filename: cython.str = path + feat.team_point[0] + "-kick_distribution"
extension: list = [".eps", ".pdf", ".png", ".svg"]
for e in extension:
plt.savefig(filename + e,
dpi=300,
bbox_inches="tight",
transparent=True)
plt.show()
def plot_graph(self, title=None, file='graph.png'):
from matplotlib.font_manager import _rebuild
_rebuild()
font_path = ''
if platform.system() is 'Windows':
# Window의 경우 폰트 경로
font_path = 'C:/Windows/Fonts/malgun.ttf'
elif platform.system() is 'Darwin':
# for Mac
font_path = '/Library/Fonts/AppleGothic.ttf'
font_name = fm.FontProperties(fname=font_path).get_name()
plt.rc('font', family=font_name)
plt.rc('axes', unicode_minus=False)
# 그래프에서 마이너스 폰트 깨지는 문제에 대한 대처
mpl.rcParams['axes.unicode_minus'] = False
#print('버전: ', mpl.__version__)
#print('설치 위치: ', mpl.__file__)
#print('설정 위치: ', mpl.get_configdir())
#print('캐시 위치: ', mpl.get_cachedir())
# size, family
print('# 설정 되어있는 폰트 사이즈')
print(plt.rcParams['font.size'])
print('# 설정 되어있는 폰트 글꼴')
print(plt.rcParams['font.family'])
fig = plt.figure(figsize=(8, 8))
pos = self.centrality_layout()
"""Conveniently summarize graph visually"""
# config parameters
edge_min_width = 3
edge_max_width = 12
label_font = 18
node_font = 22
node_alpha = 0.4
edge_alpha = 0.55
edge_cmap = plt.cm.Spectral
# Create figure
if fig is None:
fig, ax = plt.subplots()
else:
ax = fig.add_subplot(111)
fig.subplots_adjust(0, 0, 1)
font = FT2Font(font_path)
# Plot nodes with size according to count
sizes = []
degrees = []
for n, d in self.G.nodes(data=True):
sizes.append(d['count'])
degrees.append(self.G.degree(n))
sizes = self.rescale_arr(np.array(sizes, dtype=float), 100, 1000)
# Compute layout and label edges according to weight
pos = nx.spring_layout(self.G) if pos is None else pos
labels = {}
width = []
for n1, n2, d in self.G.edges(data=True):
w = d['weight']
labels[n1, n2] = w
width.append(w)
width = self.rescale_arr(np.array(width, dtype=float), edge_min_width,
edge_max_width)
# Draw
nx.draw_networkx_nodes(self.G,
pos,
node_size=sizes,
node_color=degrees,
alpha=node_alpha)
nx.draw_networkx_edges(self.G,
pos,
width=width,
edge_color=width,
edge_cmap=edge_cmap,
alpha=edge_alpha)
nx.draw_networkx_edge_labels(self.G,
pos,
edge_labels=labels,
font_size=label_font)
nx.draw_networkx_labels(self.G,
pos,
font_size=node_font,
font_family=font_name,
font_weight='bold')
if title is not None:
ax.set_title(title, fontsize=label_font)
ax.set_xticks([])
ax.set_yticks([])
# Mark centrality axes
kw = dict(color='k', linestyle='-')
cross = [ax.axhline(0, **kw), ax.axvline(self.rad0, **kw)]
[l.set_zorder(0) for l in cross]
#plt.show()
plt.savefig(file)
def main(argv):
config = read_parser(argv, Inputs, InputsOpt_Defaults)
if config['mode'] == '1x4_two_scale_plot':
from matplotlib import font_manager
del font_manager.weight_dict['roman']
font_manager._rebuild()
plt.rcParams['font.family'] = 'Times New Roman'
print('Select Features files')
root = Tk()
root.withdraw()
root.update()
Filepaths = filedialog.askopenfilenames()
root.destroy()
print('Select Feature Condition...')
root = Tk()
root.withdraw()
root.update()
filepath2 = filedialog.askopenfilename()
root.destroy()
mydict2 = pd.read_excel(filepath2, sheetname=config['sheet'])
mydict2 = mydict2.to_dict(orient='list')
Feature2 = np.array(mydict2[config['feature_cond']])
index_nonan = []
count = 0
for element in Feature2:
if not isnan(element):
index_nonan.append(count)
count +=1
Feature2 = Feature2[index_nonan]
if config['feature_cond'] == 'n':
Feature2 = Feature2/117.18
fig, ax = plt.subplots(nrows=1, ncols=2)
# ax2 = ax.twinx()
# count = 0
titles = ['AE-1', 'AE-2', 'AE-3', 'AE-4']
features = ['MAX', 'RMS']
fontsize_big = 18
fontsize_little = 17
fontsize_legend = 14
for k in range(2):
count = 0
for filepath in Filepaths:
mydict = pd.read_excel(filepath, sheetname=config['sheet'])
mydict = mydict.to_dict(orient='list')
Feature = mydict[features[k]]
Feature = 1000*np.array(Feature)/141.3
if config['n_mov_avg'] != 0:
Feature = movil_avg(Feature, config['n_mov_avg'])
time = np.array([i*10. for i in range(len(Feature))])/60.
time2 = time[index_nonan]
ax[k].plot(time, Feature, label=titles[count])
ax[k].set_xlabel('Time [min]', fontsize=fontsize_big)
if features[k] == 'RMS':
ax[k].set_ylabel('RMS value [mV]', color='k', fontsize=fontsize_big)
ax[k].set_ylim(bottom=0, top=0.8)
else:
ax[k].set_ylabel('Maximum value [mV]', color='k', fontsize=fontsize_big)
ax[k].set_ylim(bottom=0, top=40)
ax[k].tick_params('y', colors='k')
ax[k].legend(fontsize=fontsize_legend, loc='best')
ax[k].tick_params(axis='both', labelsize=fontsize_little)
ax[k].set_xlim(left=0, right=60)
count += 1
ax2 = ax[k].twinx()
ax2.set_xlim(left=0, right=60)
# ax2.set_xticks([0, 2, 4, 6, 8, 10, 12, 14])
if config['feature_cond'] == 'T':
ax2.set_ylim(bottom=25, top=55)
ax2.set_ylabel('Temperature [°C]', color='m', fontsize=fontsize_big)
elif config['feature_cond'] == 'n':
ax2.set_ylabel('Rotational speed [CPM]', color='m', fontsize=fontsize_big)
# Feature2 = Feature2/117.18
ax2.set_ylim(bottom=0, top=15)
ax2.set_yticks([0, 3, 6, 9, 12, 15])
elif config['feature_cond'] == 'M':
ax2.set_ylabel('Torque [Nm]', color='m', fontsize=fontsize_big)
# Feature2 = Feature2/117.18
ax2.set_ylim(bottom=0, top=18)
ax2.set_yticks([0, 3, 6, 9, 12, 15, 18])
# ax2.set_yticks([0, 3, 6, 9, 12, 15])
# ax2.set_ylabel('Rotational speed [CPM]', color='r', fontsize=15)
ax2.plot(time2, Feature2, 'sm')
ax2.tick_params('y', colors='m')
ax2.tick_params(axis='both', labelsize=fontsize_little)
fig.set_size_inches(12, 4.5)
plt.subplots_adjust(wspace=0.4, left=0.065, right=0.935, bottom=0.15, top=0.94)
# plt.tight_layout()
plt.show()
elif config['mode'] == 'calculate_correlation':
print('Select Features files')
root = Tk()
root.withdraw()
root.update()
Filepaths = filedialog.askopenfilenames()
root.destroy()
print('Select Feature Condition...')
root = Tk()
root.withdraw()
root.update()
filepath2 = filedialog.askopenfilename()
root.destroy()
mydict2 = pd.read_excel(filepath2, sheetname=config['sheet'])
mydict2 = mydict2.to_dict(orient='list')
Feature2 = np.array(mydict2[config['feature_cond']])
index_nonan = []
count = 0
for element in Feature2:
if not isnan(element):
index_nonan.append(count)
count +=1
Feature2 = Feature2[index_nonan]
if config['feature_cond'] == 'n':
Feature2 = Feature2/117.18
# ax2 = ax.twinx()
mydict2 = {}
count = 0
titles = ['AE-1', 'AE-2', 'AE-3', 'AE-4']
features = ['MAX', 'RMS']
for k in range(2):
vec = []
for filepath in Filepaths:
mydict = pd.read_excel(filepath, sheetname=config['sheet'])
mydict = mydict.to_dict(orient='list')
Feature = mydict[features[k]]
Feature = 1000*np.array(Feature)/141.3
if config['n_mov_avg'] != 0:
Feature = movil_avg(Feature, config['n_mov_avg'])
time = np.array([i*10. for i in range(len(Feature))])/60.
time2 = time[index_nonan]
Feature2int = np.interp(time, time2, Feature2)
# plt.plot(time, Feature, 'b', time, Feature2int, 'or')
# plt.show()
print(titles[count], ' = ' , np.corrcoef(Feature, Feature2int)[0][1])
vec.append(np.corrcoef(Feature, Feature2int)[0][1])
mydict2[features[k]] = vec
print(mydict2)
DataFr = pd.DataFrame(data=mydict2, index=titles)
writer = pd.ExcelWriter(config['name'] + '.xlsx')
DataFr.to_excel(writer, sheet_name='Correlations')
print('Result in Excel table')
else:
print('unknown mode')
sys.exit()
return
from pandas import DataFrame
from pandas import Series
from pandas.core.index import MultiIndex
from pandas.core.indexing import convert_to_index_sliceable
from performanceanalytics.charts.performance_summary import create_performance_summary
from .columns import *
from .outcomes import *
from ..io.downloader import download_latest_data
from ..util.checker import not_empty
import dropbox
import io
# Hangul font setting
# noinspection PyProtectedMember
font_manager._rebuild()
if platform.system() == 'Windows':
font_name = font_manager.FontProperties(fname='c:/Windows/Fonts/malgun.ttf').get_name()
elif platform.system() == 'Darwin': # OS X
font_name = font_manager.FontProperties(fname='/Library/Fonts/AppleGothic.ttf').get_name()
else: # Linux
fname = '/usr/share/fonts/truetype/nanum/NanumGothicOTF.ttf'
if not os.path.isfile(fname):
raise ResourceWarning("Please install NanumGothicOTF.ttf for plotting Hangul.")
font_name = font_manager.FontProperties(fname=fname).get_name()
rc('font', family=font_name)
# for fix broken Minus sign
matplotlib.rcParams['axes.unicode_minus'] = False
PERCENTAGE = 'percentage'
def main(argv):
config = read_parser(argv, Inputs, InputsOpt_Defaults)
if config['mode'] == 'plot_from_mat':
print('Select MAT from DATA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
data = load_signal(filepath, channel=config['channel'])
# data = data*(1000/70.8)*(1000/70.8)*1000*1000
# data = data*(1000/141.25)*(1000/141.25)*1000*1000
print('Select MAT from ALPHA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
alpha = load_signal(filepath, channel='alpha')
print('Select MAT from FREQ')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
f = load_signal(filepath, channel='f')
data = np.reshape(data, (int(len(f) * 2), len(alpha)))
data = data[0:int(data.shape[0] / 2), :]
data = np.absolute(data)
data[:, 0] = 0.
fig = plt.figure()
if config['projection'] == 'surface':
ax = fig.add_subplot(111, projection='3d')
alpha, f = np.meshgrid(alpha, f)
ax.plot_surface(alpha, f / 1000, data, cmap='plasma')
# ax.plot_wireframe(alpha, f/1000, data, color='black')
# ax.plot_trisurf(alpha, f/1000, data, cmap='viridis')
# elif config['projection'] == 'mesh':
# ax = fig.add_subplot(111)
# dog = ax.pcolormesh(alpha, f/1000., np.log(data), cmap='inferno', vmax=None)
# cbar = fig.colorbar(dog, ax=ax)
# cbar.set_label('SCoh [-]', fontsize=13)
# cbar.ax.tick_params(axis='both', labelsize=12)
# # cbar.ax.set_ylim(bottom=0.000001, top=20)
elif config['projection'] == 'mesh':
ax = fig.add_subplot(111)
# vmax = max_cspectrum(data)
# dog = ax.pcolormesh(alpha, f/1000, 10*np.log10(data), cmap='inferno', vmax=None)
dog = ax.pcolormesh(alpha,
f / 1000,
data,
cmap='inferno',
vmax=None)
cbar = fig.colorbar(dog, ax=ax)
# cbar.set_label('SCD [mV$^{2}$]', fontsize=13)
cbar.set_label('SCoh [-]', fontsize=13)
cbar.ax.tick_params(axis='both', labelsize=12)
# cbar.ax.set_ylim(bottom=0.000001, top=20)
elif config['projection'] == 'save_map':
extent_ = [0, np.max(alpha), 0, 500]
# ax.contourf(cwtmatr, extent=extent_)
# ax.set_ylim(bottom=0 , top=500000)
# plt.show()
mydict = {'map': data, 'extent': extent_}
save_pickle('CycloDensity_simulated_AE_mod_fault_frec.pkl', mydict)
sys.exit()
ax.set_xlabel('Cyclic Frequency [Hz]', fontsize=13)
ax.set_ylabel('Frequency [kHz]', fontsize=13)
ax.tick_params(axis='both', labelsize=12)
ax.set_title('S-1: AE-3', fontsize=12)
# ax.set_xlim(left=0, right=50)
# ax.set_ylim(bottom=0, top=500)
plt.show()
elif config['mode'] == 'plot_from_mat_cyclic':
print('Select MAT from DATA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
data = load_signal(filepath, channel=config['channel'])
# data = data*(1000/70.8)*(1000/70.8)*1000*1000
# data = data*(1000/141.25)*(1000/141.25)*1000*1000
print('Select MAT from ALPHA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
alpha = load_signal(filepath, channel='alpha')
print('Select MAT from FREQ')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
f = load_signal(filepath, channel='f')
data = np.reshape(data, (int(len(f) * 2), len(alpha)))
data = data[0:int(data.shape[0] / 2), :]
data = np.absolute(data)
sum = np.zeros(len(data[:, 0])) #sum len 100
for i in range(len(data[0, :])): #i itera de 1 a 100
sum += data[:, i] #data suma a len 100
plt.plot(f / 1000, sum)
plt.show()
sum = np.zeros(len(data[0, :])) #sum len 100
for i in range(len(data[:, 0])): #i itera de 1 a 100
sum += data[i, :] #data suma a len 100
plt.plot(alpha, sum)
plt.show()
sys.exit()
elif config['mode'] == 'avg_plot_from_mat_cyclic':
for i in range(5):
print('Select MAT from DATA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
data = load_signal(filepath, channel=config['channel'])
# data = data*(1000/70.8)*(1000/70.8)*1000*1000
# data = data*(1000/141.25)*(1000/141.25)*1000*1000
print('Select MAT from ALPHA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
alpha = load_signal(filepath, channel='alpha')
print('Select MAT from FREQ')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
f = load_signal(filepath, channel='f')
data = np.reshape(data, (int(len(f) * 2), len(alpha)))
data = data[0:int(data.shape[0] / 2), :]
data = np.absolute(data)
sum = np.zeros(len(data[:, 0])) #sum len 100
for i in range(len(data[0, :])): #i itera de 1 a 100
sum += data[:, i] #data suma a len 100
plt.plot(f / 1000, sum)
plt.show()
sum = np.zeros(len(data[0, :])) #sum len 100
for i in range(len(data[:, 0])): #i itera de 1 a 100
sum += data[i, :] #data suma a len 100
plt.plot(alpha, sum)
plt.show()
sys.exit()
elif config['mode'] == 'comp_plot_from_mat':
print('Select MAT from DATA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
data = load_signal(filepath, channel=config['channel'])
print('Select MAT from ALPHA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
alpha = load_signal(filepath, channel='alpha')
print('Select MAT from FREQ')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
f = load_signal(filepath, channel='f')
data = np.reshape(data, (int(len(f) * 2), len(alpha)))
data = data[0:int(data.shape[0] / 2), :]
data = np.absolute(data)
select_alpha = 4.3 - 2
da = alpha[1] - alpha[0]
result_alpha_idx = int(round(select_alpha / da))
result_alpha = alpha[result_alpha_idx]
plt.plot(f, data[:, result_alpha_idx])
plt.show()
print(result_alpha)
print(alpha)
sys.exit()
fig = plt.figure()
ax = fig.add_subplot(111, projection='3d')
alpha, f = np.meshgrid(alpha, f)
ax.plot_surface(alpha, f, data, cmap='plasma')
plt.show()
elif config['mode'] == 'multi_comp_plot_from_mat':
prom = 1
for i in range(prom):
print('Select MAT from DATA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
data = load_signal(filepath, channel=config['channel'])
# data = data*(1000/70.8)*(1000/70.8)*1000*1000
# data = data*(1000/141.25)*(1000/141.25)*1000*1000
# data = 10*np.log(data)
print('Select MAT from ALPHA')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
alpha = load_signal(filepath, channel='alpha')
print('Select MAT from FREQ')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
f = load_signal(filepath, channel='f')
data = np.reshape(data, (int(len(f) * 2), len(alpha)))
data = data[0:int(data.shape[0] / 2), :]
data = np.absolute(data)
#++++++++TEST RIG
# select_alphas = np.array([4.333, 4.333*2, 4.333*3, 4.333*4, 4.333*5, 4.333*6, 4.333*7, 4.333*8, 4.333*9, 4.333*10])
# select_alphas = np.array([312., 624., 936., 1248., 1560.])
# select_alphas = np.array([307.67, 316.33, 303.34, 320.66, 294.68, 329.32, 290.35, 333.65] + [619.67, 628.33, 615.34, 632.66, 606.68, 641.32, 602.35, 645.65] + [931.67, 940.33, 927.34, 944.66, 918.68, 953.32, 914.35, 957.65] + [1243.67, 1252.33, 1239.34, 1256.66, 1230.68, 1265.32, 1226.35, 1269.35] + [1555.67, 1564.33, 1551.34, 1568.66, 1542.68, 1577.32, 1538.35, 1581.65])*1.023077
# select_alphas = [4.333, 4.333*2, 4.333*3, 4.333*4, 4.333*5, 4.333*6, 4.333*7, 4.333*8, 4.333*9, 4.333*10]
# select_alphas += [312., 624., 936., 1248., 1560.]
# select_alphas += [307.67, 316.33, 303.34, 320.66, 294.68, 329.32, 290.35, 333.65] + [619.67, 628.33, 615.34, 632.66, 606.68, 641.32, 602.35, 645.65] + [931.67, 940.33, 927.34, 944.66, 918.68, 953.32, 914.35, 957.65] + [1243.67, 1252.33, 1239.34, 1256.66, 1230.68, 1265.32, 1226.35, 1269.35] + [1555.67, 1564.33, 1551.34, 1568.66, 1542.68, 1577.32, 1538.35, 1581.65]
# select_alphas = np.array(select_alphas)
# *1.023077
# #++++++++CWD DAMAGE
# # select_alphas = [35.8875, 71.775, 107.6625, 143.55, 179.4375] #fg
# select_alphas = [0.3625, 0.725, 1.0875, 1.45, 1.8125, 2.175, 2.5375, 2.9, 3.2625, 3.625] #ffr
# # select_alphas = [0.3625] #ffr
# #++++++++TURM Schottland DAMAGE
# select_alphas = [3.872, 7.744, 11.615, 15.487, 19.359, 23.231, 27.103, 30.974, 34.847, 38.718] #ffr
#++++++++BOCHUM NO DAMAGE
select_alphas = [
3.481, 6.961, 10.442, 13.923, 17.403, 20.884, 24.365, 27.485,
31.326, 34.806
] #ffr
da = alpha[1] - alpha[0]
result_alpha_idxes = []
if i == 0:
PromsComps = np.zeros(len(data[:, 0]))
Comps = np.zeros(len(data[:, 0]))
for select_alpha in select_alphas:
result_alpha_idx = int(round(select_alpha / da))
result_alpha_idxes.append(result_alpha_idx)
print(alpha[result_alpha_idx])
Comps = Comps + data[:, result_alpha_idx]
PromsComps = PromsComps + Comps
PromsComps = PromsComps / prom
fig, ax = plt.subplots()
ax.plot(f / 1000., PromsComps)
ax.set_xlabel('Frequency [kHz]', fontsize=13)
# name_ylabel = r'$\Sigma_{i=1}^{10}$ SCD $i f_{c}$ [$\mu$V$^{2}$]'
name_ylabel = r'$\Sigma_{i=1}^{10}$ SCoh $i f_{c}$ [-]'
ax.set_ylabel(name_ylabel, fontsize=13)
ax.tick_params(axis='both', labelsize=12)
ax.set_title('B-3: AE-3', fontsize=12)
plt.show()
if config['save'] == 'ON':
# #TEST-RIG
# save_pickle('MessQ_MeanDensity_5h_sb_f_g_c.pkl', PromsComps)
# save_pickle('MessQ_fDensity_5h_sb_f_g_c.pkl', f)
# #CWD-Damage
# save_pickle('CWD_Damage_MeanCoherence_10fft.pkl', PromsComps)
# save_pickle('CWD_Damage_fCoherence_10ffr.pkl', f)
#Schottland-Damage
save_pickle('Schottland_Damage_MeanCoherence_10ffr.pkl',
PromsComps)
save_pickle('Schottland_Damage_fCoherence_10ffr.pkl', f)
elif config['mode'] == 'plot_mean':
print('Select FREQ')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
f = read_pickle(filepath)
print('Select MEAN DATA')
root = Tk()
root.withdraw()
root.update()
Filepaths = filedialog.askopenfilenames()
root.destroy()
data = read_pickle(filepath)
names = ['MessA', 'MessG', 'MessI', 'MessL', 'MessN', 'MessQ']
signals = {}
for name in names:
for filepath in Filepaths:
if filepath.find(name) != -1:
if name not in signals.keys():
signals[name] = read_pickle(filepath)
font_big = 17
font_little = 15
font_label = 13
from matplotlib import font_manager
del font_manager.weight_dict['roman']
font_manager._rebuild()
plt.rcParams['font.family'] = 'Times New Roman'
fig, ax = plt.subplots(ncols=3, nrows=2, sharex=True, sharey=True)
plt.subplots_adjust(wspace=0.275,
left=0.065,
right=0.98,
bottom=0.125,
top=0.92,
hspace=0.55)
fig.set_size_inches(14.2, 6.2)
ax[0][0].plot(f / 1000., signals['MessA'])
ax[0][1].plot(f / 1000., signals['MessG'])
ax[0][2].plot(f / 1000., signals['MessI'])
ax[1][0].plot(f / 1000., signals['MessL'])
ax[1][1].plot(f / 1000., signals['MessN'])
ax[1][2].plot(f / 1000., signals['MessQ'])
name_ylabel = r'$\Sigma$ SCD [-]'
ax[0][0].set_ylabel(name_ylabel, fontsize=font_big)
ax[0][1].set_ylabel(name_ylabel, fontsize=font_big)
ax[0][2].set_ylabel(name_ylabel, fontsize=font_big)
ax[1][0].set_ylabel(name_ylabel, fontsize=font_big)
ax[1][1].set_ylabel(name_ylabel, fontsize=font_big)
ax[1][2].set_ylabel(name_ylabel, fontsize=font_big)
name_xlabel = 'Frequency [kHz]'
ax[0][0].set_xlabel(name_xlabel, fontsize=font_big)
ax[0][1].set_xlabel(name_xlabel, fontsize=font_big)
ax[0][2].set_xlabel(name_xlabel, fontsize=font_big)
ax[1][0].set_xlabel(name_xlabel, fontsize=font_big)
ax[1][1].set_xlabel(name_xlabel, fontsize=font_big)
ax[1][2].set_xlabel(name_xlabel, fontsize=font_big)
# ax[0][0].legend(fontsize=font_label, ncol=2, loc='best')
# ax[0][1].legend(fontsize=font_label, ncol=2, loc='best')
# ax[0][2].legend(fontsize=font_label, ncol=2, loc='best')
# ax[1][0].legend(fontsize=font_label, ncol=2, loc='best')
# ax[1][1].legend(fontsize=font_label, ncol=2, loc='best')
# ax[1][2].legend(fontsize=font_label, ncol=2, loc='best')
plt.rcParams['mathtext.fontset'] = 'cm'
ax[0][0].set_title('$1^{st}$ MC', fontsize=font_big)
ax[0][1].set_title('$2^{nd}$ MC', fontsize=font_big)
ax[0][2].set_title('$4^{th}$ MC', fontsize=font_big)
ax[1][0].set_title('$7^{th}$ MC', fontsize=font_big)
ax[1][1].set_title('$9^{th}$ MC', fontsize=font_big)
ax[1][2].set_title('$10^{th}$ MC', fontsize=font_big)
ax[0][0].set_xlim(left=0, right=500)
ax[0][1].set_xlim(left=0, right=500)
ax[0][2].set_xlim(left=0, right=500)
ax[1][0].set_xlim(left=0, right=500)
ax[1][1].set_xlim(left=0, right=500)
ax[1][2].set_xlim(left=0, right=500)
valtop = 0.4
ax[0][0].set_ylim(bottom=0, top=valtop)
ax[0][1].set_ylim(bottom=0, top=valtop)
ax[0][2].set_ylim(bottom=0, top=valtop)
ax[1][0].set_ylim(bottom=0, top=valtop)
ax[1][1].set_ylim(bottom=0, top=valtop)
ax[1][2].set_ylim(bottom=0, top=valtop)
# ax[0][0].set_yticks([0, 100, 200, 300, 400, 500])
# ax[0][1].set_yticks([0, 100, 200, 300, 400, 500])
# ax[0][2].set_yticks([0, 100, 200, 300, 400, 500])
# ax[1][0].set_yticks([0, 100, 200, 300, 400, 500])
# ax[1][1].set_yticks([0, 100, 200, 300, 400, 500])
# ax[1][2].set_yticks([0, 100, 200, 300, 400, 500])
ax[0][0].tick_params(axis='both', labelsize=font_little)
ax[0][1].tick_params(axis='both', labelsize=font_little)
ax[0][2].tick_params(axis='both', labelsize=font_little)
ax[1][0].tick_params(axis='both', labelsize=font_little)
ax[1][1].tick_params(axis='both', labelsize=font_little)
ax[1][2].tick_params(axis='both', labelsize=font_little)
for ax_it in ax.flatten():
for tk in ax_it.get_yticklabels():
tk.set_visible(True)
for tk in ax_it.get_xticklabels():
tk.set_visible(True)
ax_it.yaxis.offsetText.set_visible(True)
plt.show()
elif config['mode'] == 'plot_mean_multi':
print('Select FREQ')
root = Tk()
root.withdraw()
root.update()
filepath = filedialog.askopenfilename()
root.destroy()
f = read_pickle(filepath)
print('Select MEAN DATA')
root = Tk()
root.withdraw()
root.update()
Filepaths = filedialog.askopenfilenames()
root.destroy()
data = read_pickle(filepath)
names = ['MessA', 'MessG', 'MessI', 'MessL', 'MessN', 'MessQ']
harmonics = ['10h_f_f_r', '5h_f_g', '5h_sb_f_g_c']
signals = {}
for name in names:
for harmonic in harmonics:
for filepath in Filepaths:
if filepath.find(name) != -1 and filepath.find(
harmonic) != -1:
if name not in signals.keys():
signals[name + '_' +
harmonic] = read_pickle(filepath)
font_big = 17 + 3
font_little = 15 + 3
font_label = 13 + 3
font_caption = 23 + 3
from matplotlib import font_manager
del font_manager.weight_dict['roman']
font_manager._rebuild()
plt.rcParams['font.family'] = 'Times New Roman'
fig, ax = plt.subplots(ncols=3, nrows=2, sharex=True, sharey=True)
# plt.subplots_adjust(wspace=0.275, left=0.065, right=0.98, bottom=0.125, top=0.92, hspace=0.55)
plt.subplots_adjust(wspace=0.32,
left=0.07,
right=0.985,
bottom=0.135,
top=0.94,
hspace=0.52)
# fig.set_size_inches(14.2, 6.2)
fig.set_size_inches(14.2, 7)
fig.text(0.053 - 0.015, 0.04, '(d)', fontsize=font_caption)
fig.text(0.385 - 0.015, 0.04, '(e)', fontsize=font_caption)
fig.text(0.717 - 0.015, 0.04, '(f)', fontsize=font_caption)
fig.text(0.053 - 0.015, 0.528, '(a)', fontsize=font_caption)
fig.text(0.385 - 0.015, 0.528, '(b)', fontsize=font_caption)
fig.text(0.717 - 0.015, 0.528, '(c)', fontsize=font_caption)
ax[0][0].plot(f / 1000., signals['MessA_10h_f_f_r'], label='$n f_{c}$')
ax[0][0].plot(f / 1000., signals['MessA_5h_f_g'], label='$p f_{m}$')
ax[0][0].plot(f / 1000.,
signals['MessA_5h_sb_f_g_c'],
label='$p f_{m} \pm s f_{c}$')
ax[0][1].plot(f / 1000., signals['MessG_10h_f_f_r'], label='$n f_{c}$')
ax[0][1].plot(f / 1000., signals['MessG_5h_f_g'], label='$p f_{m}$')
ax[0][1].plot(f / 1000.,
signals['MessG_5h_sb_f_g_c'],
label='$p f_{m} \pm s f_{c}$')
ax[0][2].plot(f / 1000., signals['MessI_10h_f_f_r'], label='$n f_{c}$')
ax[0][2].plot(f / 1000., signals['MessI_5h_f_g'], label='$p f_{m}$')
ax[0][2].plot(f / 1000.,
signals['MessI_5h_sb_f_g_c'],
label='$p f_{m} \pm s f_{c}$')
ax[1][0].plot(f / 1000., signals['MessL_10h_f_f_r'], label='$n f_{c}$')
ax[1][0].plot(f / 1000., signals['MessL_5h_f_g'], label='$p f_{m}$')
ax[1][0].plot(f / 1000.,
signals['MessL_5h_sb_f_g_c'],
label='$p f_{m} \pm s f_{c}$')
ax[1][1].plot(f / 1000., signals['MessN_10h_f_f_r'], label='$n f_{c}$')
ax[1][1].plot(f / 1000., signals['MessN_5h_f_g'], label='$p f_{m}$')
ax[1][1].plot(f / 1000.,
signals['MessN_5h_sb_f_g_c'],
label='$p f_{m} \pm s f_{c}$')
ax[1][2].plot(f / 1000., signals['MessQ_10h_f_f_r'], label='$n f_{c}$')
ax[1][2].plot(f / 1000., signals['MessQ_5h_f_g'], label='$p f_{m}$')
ax[1][2].plot(f / 1000.,
signals['MessQ_5h_sb_f_g_c'],
label='$p f_{m} \pm s f_{c}$')
# name_ylabel = r'$\Sigma$ SCD [mV$^{2}$]'
name_ylabel = r'$\Sigma$ SC [-]'
ax[0][0].set_ylabel(name_ylabel, fontsize=font_big)
ax[0][1].set_ylabel(name_ylabel, fontsize=font_big)
ax[0][2].set_ylabel(name_ylabel, fontsize=font_big)
ax[1][0].set_ylabel(name_ylabel, fontsize=font_big)
ax[1][1].set_ylabel(name_ylabel, fontsize=font_big)
ax[1][2].set_ylabel(name_ylabel, fontsize=font_big)
name_xlabel = 'Frequency [kHz]'
ax[0][0].set_xlabel(name_xlabel, fontsize=font_big)
ax[0][1].set_xlabel(name_xlabel, fontsize=font_big)
ax[0][2].set_xlabel(name_xlabel, fontsize=font_big)
ax[1][0].set_xlabel(name_xlabel, fontsize=font_big)
ax[1][1].set_xlabel(name_xlabel, fontsize=font_big)
ax[1][2].set_xlabel(name_xlabel, fontsize=font_big)
ax[0][0].legend(fontsize=font_label,
ncol=1,
loc='best',
labelspacing=0.2,
handletextpad=0.2)
ax[0][1].legend(fontsize=font_label,
ncol=1,
loc='best',
labelspacing=0.2,
handletextpad=0.2)
ax[0][2].legend(fontsize=font_label,
ncol=1,
loc='best',
labelspacing=0.2,
handletextpad=0.2)
ax[1][0].legend(fontsize=font_label,
ncol=1,
loc='best',
labelspacing=0.2,
handletextpad=0.2)
ax[1][1].legend(fontsize=font_label,
ncol=1,
loc='best',
labelspacing=0.2,
handletextpad=0.2)
ax[1][2].legend(fontsize=font_label,
ncol=1,
loc='best',
labelspacing=0.2,
handletextpad=0.2)
plt.rcParams['mathtext.fontset'] = 'cm'
# ax[0][0].set_title('$1^{st}$ MC', fontsize=font_big)
# ax[0][1].set_title('$2^{nd}$ MC', fontsize=font_big)
# ax[0][2].set_title('$4^{th}$ MC', fontsize=font_big)
# ax[1][0].set_title('$7^{th}$ MC', fontsize=font_big)
# ax[1][1].set_title('$9^{th}$ MC', fontsize=font_big)
# ax[1][2].set_title('$10^{th}$ MC', fontsize=font_big)
ax[0][0].set_title('MC 1', fontsize=font_big)
ax[0][1].set_title('MC 2', fontsize=font_big)
ax[0][2].set_title('MC 4', fontsize=font_big)
ax[1][0].set_title('MC 7', fontsize=font_big)
ax[1][1].set_title('MC 9', fontsize=font_big)
ax[1][2].set_title('MC 10', fontsize=font_big)
ax[0][0].set_xlim(left=0, right=500)
ax[0][1].set_xlim(left=0, right=500)
ax[0][2].set_xlim(left=0, right=500)
ax[1][0].set_xlim(left=0, right=500)
ax[1][1].set_xlim(left=0, right=500)
ax[1][2].set_xlim(left=0, right=500)
# valtop = 0.15
valtop = 1.
ax[0][0].set_ylim(bottom=0, top=valtop)
ax[0][1].set_ylim(bottom=0, top=valtop)
ax[0][2].set_ylim(bottom=0, top=valtop)
ax[1][0].set_ylim(bottom=0, top=valtop)
ax[1][1].set_ylim(bottom=0, top=valtop)
ax[1][2].set_ylim(bottom=0, top=valtop)
ax[0][0].set_xticks([0, 100, 200, 300, 400, 500])
ax[0][1].set_xticks([0, 100, 200, 300, 400, 500])
ax[0][2].set_xticks([0, 100, 200, 300, 400, 500])
ax[1][0].set_xticks([0, 100, 200, 300, 400, 500])
ax[1][1].set_xticks([0, 100, 200, 300, 400, 500])
ax[1][2].set_xticks([0, 100, 200, 300, 400, 500])
ax[0][0].tick_params(axis='both', labelsize=font_little)
ax[0][1].tick_params(axis='both', labelsize=font_little)
ax[0][2].tick_params(axis='both', labelsize=font_little)
ax[1][0].tick_params(axis='both', labelsize=font_little)
ax[1][1].tick_params(axis='both', labelsize=font_little)
ax[1][2].tick_params(axis='both', labelsize=font_little)
ax[0][0].grid(axis='both')
ax[0][1].grid(axis='both')
ax[0][2].grid(axis='both')
ax[1][0].grid(axis='both')
ax[1][1].grid(axis='both')
ax[1][2].grid(axis='both')
for ax_it in ax.flatten():
for tk in ax_it.get_yticklabels():
tk.set_visible(True)
for tk in ax_it.get_xticklabels():
tk.set_visible(True)
ax_it.yaxis.offsetText.set_visible(True)
# plt.show()
name = 'TestBench_Cyclo_SC_Each_AE_EnvFftComps'
path_1 = 'C:\\Felix\\29_THESIS\\MODEL_A\\Chapter3_Test_Bench\\03_Figures\\'
path_2 = 'C:\\Felix\\29_THESIS\\MODEL_A\\LATEX_Diss_FLn\\bilder\\Figures_Test_Bench\\'
path_1b = path_1 + name + '.svg'
path_2b = path_2 + name + '.pdf'
plt.savefig(path_1b)
plt.savefig(path_2b)
else:
print('wrong mode')
return
def printSequence(sp: la_class.ServerParam, feat: la_class.Feature,
path: str) -> None:
import matplotlib
# smatplotlib.use('Agg')
from matplotlib import pyplot as plt
import matplotlib.font_manager as fm
fig = plt.figure(figsize=(10.5, 6.8))
plt.rcParams['font.family'] = 'Times New Roman'
plt.rcParams['ps.useafm'] = True
plt.rcParams['pdf.use14corefonts'] = True
plt.rcParams['text.usetex'] = True
xlim = sp.pitch_length / 2 + 5.0
ylim = sp.pitch_width / 2 + 5.0
fm._rebuild()
plt.xlim([-xlim, xlim])
plt.ylim([ylim, -ylim])
plt.tick_params(labelsize=32)
# plot soccer fields
plt.plot([sp.goal_line_l, -sp.penalty_area_x],
[-sp.penalty_area_y, -sp.penalty_area_y],
color='g',
linewidth=4)
plt.plot([sp.goal_line_l, -sp.penalty_area_x],
[sp.penalty_area_y, sp.penalty_area_y],
color='g',
linewidth=4)
plt.plot([sp.goal_line_r, sp.penalty_area_x],
[-sp.penalty_area_y, -sp.penalty_area_y],
color='g',
linewidth=4)
plt.plot([sp.goal_line_r, sp.penalty_area_x],
[sp.penalty_area_y, sp.penalty_area_y],
color='g',
linewidth=4)
plt.plot([-sp.penalty_area_x, -sp.penalty_area_x],
[sp.penalty_area_y, -sp.penalty_area_y],
color='g',
linewidth=4)
plt.plot([sp.penalty_area_x, sp.penalty_area_x],
[sp.penalty_area_y, -sp.penalty_area_y],
color='g',
linewidth=4)
plt.plot([sp.goal_line_l, sp.goal_line_r],
[-sp.pitch_width / 2, -sp.pitch_width / 2],
color='g',
linewidth=4)
plt.plot([sp.goal_line_l, sp.goal_line_r],
[sp.pitch_width / 2, sp.pitch_width / 2],
color='g',
linewidth=4)
plt.plot([sp.goal_line_l, sp.goal_line_l],
[-sp.pitch_width / 2, sp.pitch_width / 2],
color='g',
linewidth=4)
plt.plot([sp.goal_line_r, sp.goal_line_r],
[-sp.pitch_width / 2, sp.pitch_width / 2],
color='g',
linewidth=4)
plt.plot([0, 0], [-sp.pitch_width / 2, sp.pitch_width / 2],
color='g',
linewidth=4)
p = plt.Circle((0.0, 0.0), 9.0, color='g', linewidth=4, fill=False)
ax = plt.gca()
ax.add_patch(p)
for x, y, color in zip(feat.all_kick_path_x, feat.all_kick_path_y,
feat.color4plt_ks):
plt.plot(x, y, color)
filename = path + feat.team_point[0] + "-kick_sequence"
extension = [".eps", ".pdf", ".png", ".svg"]
for e in extension:
plt.savefig(filename + e,
dpi=300,
bbox_inches="tight",
transparent=True)
plt.show()
#!/usr/bin/python
#coding:utf-8
from skimage import io, transform
import numpy as np
import matplotlib.pyplot as plt
from matplotlib.font_manager import _rebuild
import tensorflow as tf
from pylab import mpl
_rebuild() # 重新创建字体索引列表
mpl.rcParams['font.sans-serif']=[u'SimHei'] # 正常显示中文标签
mpl.rcParams['axes.unicode_minus']=False # 正常显示正负号
def load_image(path):
fig = plt.figure("Centre and Resize")
img = io.imread(path)
img = img / 255.0
ax0 = fig.add_subplot(131)
ax0.set_xlabel(u'Original Picture')
ax0.imshow(img)
short_edge = min(img.shape[:2])
y = (img.shape[0] - short_edge) // 2
x = (img.shape[1] - short_edge) // 2
crop_img = img[y:y+short_edge, x:x+short_edge]
ax1 = fig.add_subplot(132)
ax1.set_xlabel(u"Centre Picture")
ax1.imshow(crop_img)
# 在terminal里画图会报错 _tkinter.TclError: no display name and no $DISPLAY environment variable
# 必须加下面来防止出错
import matplotlib
matplotlib.use('Agg')
# plt.clf() # 清理之前的内容。好像plt.savefig和plt.show()不一样，似乎有时候不会清理之前的内容
plt.savefig('filename.png', bbox_inches='tight') # use this instead of plt.show()
# bbox_inches='tight' 的效果非常赞，它可以让两边边变窄， 周围的文字留全！！！！
# 中文显示: 来自Dong Zhou, 还未自己尝试
# 需要自己下载字体 anaconda3/lib/python3.6/site-packages/matplotlib/mpl-data/fonts/ttf
plt.rcParams['font.sans-serif'] = 'SimHei'
plt.rcParams['axes.unicode_minus'] = False
# 还得删除缓存：https://www.zhihu.com/question/25404709/answer/309784195
from matplotlib.font_manager import _rebuild
_rebuild() #reload一下
# use seaborn
import seaborn as sns; sns.set(color_codes=True)
# 如果pandas希望在jupyter里display所有的列，那么用下面的代码
# https://stackoverflow.com/questions/11361985/output-data-from-all-columns-in-a-dataframe-in-pandas
pd.set_option('display.max_columns', None)
# 如果希望显示pandas cell内部的所有内容:
# https://stackoverflow.com/questions/25351968/how-to-display-full-non-truncated-dataframe-information-in-html-when-convertin
pd.set_option('display.max_colwidth', -1)
from matplotlib.colors import LinearSegmentedColormap
from multiprocessing import Pool, sharedctypes, RawArray, RawValue
from CMIP import *
from MS.IBIS import *
from MS.LPJ import *
from MS.Biome_BGC import *
from MS.MOD17A2 import *
import pymongo
client = pymongo.MongoClient(host=['223.2.41.104:27017'])
db = client.Comparison
metricCollection = db.Metric
myfont = fm.FontProperties(fname='/home/scr/.config/matplotlib/msyh.ttf')
plt.switch_backend('Agg')
_rebuild()
sns.set()
# annual/model-avg.nc
def plotSpatial(argv):
fig, axes = plt.subplots(nrows=int(len(argv['models'])),
ncols=1,
figsize=(8, 16),
dpi=100,
sharex=True,
sharey=True)
axes = axes.flatten()
for i, model in enumerate(argv['models']):
fpath = '../data/annual/%s-avg.nc' % model
dataset = Dataset(fpath, mode='r')
# %matplotlib inline
import matplotlib as mpl
import matplotlib.pyplot as plt
import matplotlib.font_manager as fm
!apt-get update -qq
!apt-get install fonts-nanum* -qq
path = '/usr/share/fonts/truetype/nanum/NanumBarunGothic.ttf'
font_name = fm.FontProperties(fname=path, size=10).get_name()
print(font_name)
plt.rc('font', family=font_name)
fm._rebuild()
mpl.rcParams['axes.unicode_minus'] = False
# drive mount
from google.colab import drive
drive.mount('/content/drive')
# colab to pdf install
# !apt-get install texlive texlive-xetex texlive-latex-extra pandoc
# !pip install pypandoc
# colab to pdf code
# !jupyter nbconvert --to PDF '/content/drive/MyDrive/Colab Notebooks/bigdata_project_scv.ipynb'
import matplotlib.pyplot as plt
import matplotlib as mpl
import matplotlib.font_manager as fm
import numpy as np
%matplotlib inline
%config InlineBackend.figure_format = 'retina‘
mpl.rcParams['axes.unicode_minus'] = False # 하이픈(-) 폰트 깨질 경우 대비
path = '/usr/share/fonts/truetype/nanum/NanumGothicBold.ttf'
font_name = fm.FontProperties(fname=path, size=18).get_name()
plt.rc('font', family=font_name)
fm._rebuild() # 이걸 해줘야 plt.rc가 작동
plt.plot(np.random.randn(4, 8), np.random.randn(4,8), 'bo--')
plt.title('타이틀')
plt.xlabel('X 라벨')
plt.ylabel('Y 라벨')
plt.show()
from matplotlib.font_manager import _rebuild
from IPython.display import display
# plt.style.use("fivethirtyeight")
# sns.set_style({'font.sans-serif':['simhei','Arial']})
plt.rcParams['font.sans-serif'] = ['simHei'] #用来正常显示中文标签
plt.rcParams['font.size'] = 6
# plt.rcParams['axes.unicode_minus']=False #用来正常显示负号
# 导入数据
sz_rent = pd.read_csv('merge_zc.csv')
# print('data-row2====')
display(sz_rent.head(n=2))
_rebuild() #reload一下
# # 检查缺失值情况
# print('info====')
sz_rent.info()
print('describe====')
print(sz_rent.describe())
# 添加新特征均价
zf = sz_rent.copy()
zf['perPrice'] = sz_rent['price'] / sz_rent['size']
#print('PerPrice====',sz_rent['PerPrice'])
# # 重新摆放列位置
columns = [
def main_mini(lujing):
print('jinrule ', 'main_mini')
data = [i for i in open(lujing, encoding='utf-8').readlines()]
out = []
for i in data:
for j in i.split(';'):
j = j.strip("\n")
j = j.strip(' ')
out.append(j)
# print(out[:10])
# print(len(out))
print('jinrule ', 'main_mini2')
from collections import defaultdict
dict1 = defaultdict(int)
print('jinrule ', 'main_mini3')
for i in out:
dict1[i] += 1
print('jinrule ', 'main_mini4')
# print(len(dict1))
a = []
for i in dict1:
a.append((i, dict1[i]))
a = sorted(a, key=lambda x: x[1])
a = a[::-1]
print('jinrule ', 'main_mini5')
# print(a[:1000])
# 我们只处理最大前1千个词语
a = a[:
100] ####################################################################################
print(778, a)
# 计算文本相似度.
from bert_serving.client import BertClient
import numpy as np
bc = BertClient()
print(32432423473947, a)
shuju = [i[0] for i in a]
def cosine(a, b):
return a.dot(b) / (np.linalg.norm(a) * np.linalg.norm(b))
print('jinrule ', shuju)
emb = np.array(bc.encode(shuju))
# print('new',emb)
# print(['First do it', 'then do it right'], ":", cosine(emb[0], emb[1]))
# return 99999999999999
# 下面调用这个函数即可vec_route
# b里面放的是词语,词频,词向量
b = []
for i in range(len(a)):
b.append([a[i][0], a[i][1], emb[i]])
# print(b,333333333333333333333333)
# 下面进行降维
from matplotlib.font_manager import _rebuild
_rebuild() #reload一下
import numpy as np
from sklearn.manifold import TSNE
import matplotlib.pyplot as plt
import matplotlib as mpl
label = []
for i in b:
label.append(i[0])
vectors = []
for i in b:
vectors.append(i[2])
cipin = []
for i in b:
cipin.append(i[1])
print(112)
tsne = TSNE(perplexity=30,
n_components=2,
init='pca',
n_iter=5000,
method='exact')
low_dim_embs = tsne.fit_transform(
vectors) # 需要显示的词向量，一般比原有词向量文件中的词向量个数少，不然点太多，显示效果不好
print(113)
for i in range(len(b)):
b[i][2] = low_dim_embs[i]
uuuuu = str(b)
print(11233333333)
import json
#chuanshu22288.json 这个数据就是给前端用的!!!!!!!!!!!!!!!!!1
bb = b
for i in range(len(bb)):
bb[i][2] = list(bb[i][2])
print(11255555555)
import json
print(1127777777777)
# define A.class
class node:
def __init__(self, id, label, x, y, size, color):
self.id = id
self.label = label
self.x = x
self.y = y
self.size = size
self.cluster = color # cluster
##
import matplotlib.pyplot as plt
from sklearn.datasets.samples_generator import make_blobs
from sklearn.cluster import KMeans
from sklearn import metrics
print(11299999999)
#选择聚类数K=2 聚类小于=8,因为颜色就写了8个
final2 = []
for i4 in range(3, 11):
n_clusters = i4
y_pred = KMeans(n_clusters=n_clusters).fit_predict(low_dim_embs)
colorlist = [
'red', 'black', 'yellow', 'greenyellow', 'blue', 'brown', 'coral',
'cyan', 'deeppink', 'orange'
]
list1 = []
for i, j in enumerate(bb):
list1.append(
node(i, j[0], float((float(j[2][0]))), float((float(j[2][1]))),
float(j[1]**0.5), colorlist[y_pred[i]]).__dict__)
nodes = {}
# 接口:https://github.com/xukuanzhuo/xukuanzhuo.github.io/issues/8
#下面把每一个聚类里面的距离算一下.
#similarity
class edge:
def __init__(self, sourceID, targetID, size):
self.source = sourceID
self.target = targetID
self.size = size
print(112000000000)
print(bb, 'lookupu234')
list3 = []
for i in range(n_clusters):
dexlist = []
for j in range(len(y_pred)):
if y_pred[j] == i:
dexlist.append(j)
for i1 in range(len(dexlist)):
for i2 in range(i1 + 1, len(dexlist)):
left = dexlist[i1] # 得到索引.
right = dexlist[i2]
similar = cosine(np.array(bb[left][2]),
np.array(bb[right][2])) + 1
list3.append(
edge(bb[left][0], bb[right][0], similar).__dict__)
print(list3, "list3")
all3 = {"clusterNum": n_clusters, "nodes": list1, "edges": list3}
final2.append(all3)
print('final2', final2)
tmp2 = json.dumps(final2, ensure_ascii=False)
with open("chuanshuFinal.json", mode='w', encoding='utf-8') as f:
print("最终json放在chuanshuFinal.json")
f.write(tmp2)
return tmp2
