-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathutils_mpl.py
146 lines (128 loc) · 5.32 KB
/
utils_mpl.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
from matplotlib.widgets import RectangleSelector
from matplotlib.widgets import LassoSelector
import matplotlib.colors as clr
from matplotlib.path import Path
from matplotlib.patches import Rectangle, Ellipse,Polygon
import numpy as np
from matplotlib.ticker import MaxNLocator
class Highlighter(object):
def __init__(self, ax_umap, ax_year, ax_mesh, dfpapers, colors='c', method='lasso', add_text=False,index_list=None,include_nlp=False):
self.ax_umap = ax_umap
self.ax_year = ax_year
self.ax_mesh = ax_mesh
self.canvas_umap = ax_umap.figure.canvas
self.canvas_year = ax_year.figure.canvas
self.canvas_mesh = ax_mesh.figure.canvas
self.x, self.y = dfpapers['umap-x'], dfpapers['umap-y']
self.add_text = add_text
self.df_papers = dfpapers
self.df_selected = None
#self.selected_text = None
self.full_index = dfpapers.index.values
highlight_color = colors
highlight_size = 20
self.object_index = None
self.mask = np.zeros(self.x.shape[0], dtype=bool)
self.scaled_rgb = (65 / 256., 91 / 256., 169 / 256.)
self.selection_method = method
self.x0 = None
self.y0 = None
self.x1 = None
self.y1 = None
self.entities_dict = {}
self._highlight = ax_umap.scatter(
[],[], marker="o", lw=0, s=highlight_size, c=highlight_color
)
self.include_nlp = include_nlp
if include_nlp:
self.nlp = spacy.load('en_ner_bionlp13cg_md')
self.nlp.add_pipe(self.nlp.create_pipe('sentencizer'))
#if self.selected_text != None:
# self.selected_text.remove()
if self.add_text:
self.selected_text = self.ax_umap.text(
0.03,
0.97,
"%3.2f (%i events)" % (0, 0),
ha="left",
va="top",
transform=self.ax_umap.transAxes,
color="k",
fontname="Helvetica",
)
self.name_to_selector = {"lasso": LassoSelector}
selector = self.name_to_selector[method.lower()]
onselect_dict = {
LassoSelector: self._onselect_lasso,
}
self.lasso = selector(self.ax_umap, onselect_dict[selector])
# def remove_text(self):
def disconnect(self):
self.canvas_umap.mpl_disconnect(self.lasso)
def _draw_points(self):
self.object_index = self.full_index[self.mask]
self.nevents = len(self.mask)
self.nselected = np.sum(self.mask)
if self.add_text:
self.selected_text.remove()
if self.add_text:
self.selected_text = self.ax_umap.text(
0.03,
0.97,
"%3.2f (%i events)"
% (self.nselected * 100. / self.nevents, self.nselected),
ha="left",
va="top",
transform=self.ax_umap.transAxes,
color="k",
fontname="Helvetica",
)
xy = np.column_stack([self.x[self.mask], self.y[self.mask]])
self._highlight.set_offsets(xy)
self.canvas_umap.draw()
def _onselect_lasso(self, verts):
self.verts = np.array(verts)
p = Path(self.verts)
pix = np.array([self.x, self.y]).T
self.mask = p.contains_points(pix, radius=0)
self._draw_points()
self.df_selected = self.df_papers[self.mask]
print("%s selected" % self.df_selected.shape[0])
if self.df_selected.shape[0] == 0:
self.entities_dict = {}
self.ax_mesh.clear()
self.ax_year.clear()
mterms = []
for idx in self.df_selected.index:
yp = self.df_selected.loc[idx]['pubdate']
mt = self.df_selected.loc[idx]['mesh_terms']
t = self.df_selected.loc[idx]['title']
#sdoc = self.nlp(self.df_selected.loc[idx]['abstract'])
#sen = list(sdoc.sents)
mterms.extend(sorted([mti.split(":")[-1] for mti in mt.split(";")]))
print("(%s) %s" % (yp,t))
doci = self.df_selected.loc[idx]['title'] + self.df_selected.loc[idx]['abstract']
if self.include_nlp:
sdoc = nlp(doci)
#for si in sdoc.ents:
self.entities_dict[idx] = sdoc.ents
#print(" Mesh:",", ".join(mterms))
# print(" ".join(item.split()[:100]))
self.df_selected.to_csv("tmp_selected.csv")
self.mesh_counts = Counter(mterms)
self.mesh_counts_top = self.mesh_counts.most_common(50)
dfcounts = pd.DataFrame.from_dict(dict(self.mesh_counts_top), orient='index')
dfcounts.sort_values(0,inplace=True)
if dfcounts.shape[0]!=0:
dfcounts.plot(kind='barh',ax=self.ax_mesh,fontsize=6,legend=False)
xmin = np.min(self.df_selected['pubdate'])
xmax = np.max(self.df_selected['pubdate'])
self.ax_year.hist(self.df_selected['pubdate'],bins=np.arange(xmin,xmax+1)-0.5)
self.ax_year.set_xlim([xmin,xmax])
self.ax_year.xaxis.set_major_locator(MaxNLocator(integer=True))
for label in self.ax_year.get_xticklabels():
label.set_rotation(45)
label.set_fontsize(6)
label.set_ha('right')
self.canvas_year.draw()
self.canvas_mesh.draw()