def view_command():
for i in list1.tree.get_children():
list1.tree.delete(i)
list1._build_tree(c_header, pddf(columns=c_header))
df=pddf(backend.view(), columns=c_header)
for c in ['id',
'mother_id',
'father_id',
'spouse_id',
'generation',
'cluster',
'gender' ]:
df[c]=to_numeric(df[c], downcast='integer',errors='coerce')
list1._build_tree(c_header, df)
def file_save_xls():
from tkinter.filedialog import asksaveasfilename
f = asksaveasfilename(title = "Select file",filetypes = (("Excel files","*.xlsx"),("all files","*.*")), defaultextension='.png')
if f is None: # asksaveasfile return `None` if dialog closed with "cancel".
return
df=pddf(backend.view(), columns=c_header)
import io
df.to_excel(f, index = False)
def values_greater_than(my_dataframe, my_col, threshold=0):
counts = my_dataframe[my_col].value_counts()
fields_kept = []
for i in range(len(counts)):
if counts[i] > threshold:
fields_kept.append(counts.index[i])
print("Keeping:", fields_kept)
temp = pddf()
for item in fields_kept:
temp = temp.append(my_dataframe[my_dataframe[my_col] == item])
return temp
def mid_to_matrix(mid, output='nested_list'): # {"nested_list", "pandas"}
"""
Takes a midi file or stream and returns a matrix with rows representing midi events
and columns representing midinote, offset position and event duration:
| | pitch | offset | duration | velocity | channel |
| ---- | ----- | ------ | -------- | -------- | ------- |
| evt1 | | | | | |
| evt2 | | | | | |
| evt3 | | | | | |
| ... | | | | | |
"""
mid = parse_mid(mid)
if mid.type is not 0:
print("Midi file type {}. Reformat to type 0 before quantising.".format(mid.type))
return None
resolution = mid.ticks_per_beat
elapsed = 0
noteons = []
offsets = []
noteoffs = []
durations = []
for msg in mid.tracks[0]:
elapsed += msg.time
offset = elapsed / resolution
if msg.type == 'note_on':
noteons.append(msg.note)
offsets.append(offset)
if msg.type == 'note_off':
noteoffs.append(msg.note)
durations.append(offset)
if not len(noteons) == len(noteoffs):
print("Unmatcghing size. Reformat file first")
return None
else:
mnotes = []
for i in range(len(noteons)):
mnotes.append([noteons[i], offsets[i], durations[noteoffs.index(noteons[i])] - offsets[i]])
durations.pop(noteoffs.index(noteons[i]))
noteoffs.remove(noteons[i])
if output is 'nested_list':
return mnotes
elif output is 'pandas':
return pddf(mnotes, columns=['pitch', 'offset', 'duration'])
def n_most_frequent_values(my_dataframe, my_col, n_most_freq=6):
counts = my_dataframe[my_col].value_counts()
if len(counts) <= n_most_freq:
return my_dataframe
elif len(counts) > n_most_freq:
counts = counts[:n_most_freq]
print("Keeping:", counts.index)
temp = pddf()
for item in counts.index:
temp = temp.append(my_dataframe[my_dataframe[my_col] == item])
return temp
def draw_tree():
from PIL import ImageTk, Image as Image_PIL
class ResizingCanvas(Canvas):
def __init__(self,parent,**kwargs):
Canvas.__init__(self,parent,**kwargs)
self.bind("<Configure>", self.on_resize)
self.height = self.winfo_reqheight()
self.width = self.winfo_reqwidth()
# self.bind_all("<MouseWheel>", self.on_mousewheel)
def on_resize(self,event):
new_width = self.master.winfo_width()
new_height = self.master.winfo_height()
#colors
col_f='#ff4e5c'
col_m='#1c3144'
col_w='#9d7a7d'
#
top = tk.Tk()
top.iconbitmap(f'{main_f}2683232.ico')
Grid.rowconfigure(top, 0, weight=1)
Grid.columnconfigure(top, 0, weight=1)
top.geometry("900x480")
top.wm_title("Tree")
frame = tk.Frame(top)
frame.grid(row=0, column=0, sticky=N+S+E+W)
#scrollbars
xscrollbar = tk.Scrollbar(frame, orient=tk.HORIZONTAL)
xscrollbar.grid(row=1, column=0, sticky="EW")
yscrollbar = tk.Scrollbar(frame)
yscrollbar.grid(row=0, column=1, sticky="NS")
canvas = ResizingCanvas(frame, xscrollcommand=xscrollbar.set, yscrollcommand=yscrollbar.set)
canvas.grid(row=0, column=0, sticky="EWNS")
canvas.configure(bg='white')
# Image
import pydotplus as pdt
import time
df=pddf(backend_app.view(), columns=c_header)
for c in ['id',
'mother_id',
'father_id',
'spouse_id',
'generation',
'cluster',
'gender' ]:
df[c]=to_numeric(df[c], downcast='integer',errors='coerce')
min_level=int(df['generation'].min())
max_level=int(df['generation'].max())
g = pdt.Dot(graph_type='digraph', compound='true', rankdir='TB',newrank = 'true')
i=0
edges = []
nodes=[]
# create all edges
from ast import literal_eval as make_tuple
nl=max_level-min_level+1
s=[None]*(max_level+1)
for i in range(min_level,max_level+1):
s = pdt.Subgraph(i,rank='same')
for j in df.loc[df['generation']==i, 'id'].values :
s.add_node(pdt.Node(f"{df.loc[df['id']==j, 'id'].values[0]}"))
g.add_subgraph(s)
min_cl=int(df['cluster'].min())
max_cl=int(df['cluster'].max())
nl=max_cl-min_cl+1
ss=[None]*(max_cl+1)
for i in range(min_cl,max_cl+1):
ss = pdt.Cluster(f'c1_{i}', color="white",rankdir="TB")
for j in df.loc[df['cluster']==i, 'id'].values :
ss.add_node(pdt.Node(f"{df.loc[df['id']==j, 'id'].values[0]}"))
g.add_subgraph(ss)
for i in range(df.shape[0]):
m=df.loc[i, 'mother_id']
f=df.loc[i, 'father_id']
nod=pdt.Node(f"{df.loc[i, 'id']}", shape='record', color=col_m, style='rounded')
if m in df['id'].values: g.add_edge(pdt.Edge(pdt.Node(f"{df.loc[df['id']==m, 'id'].values[0]}"), nod, color=col_f))
if f in df['id'].values: g.add_edge(pdt.Edge(pdt.Node(f"{df.loc[df['id']==f, 'id'].values[0]}"), nod, color=col_m))
if notnull(df.loc[i, 'spouse_id']):
who=to_numeric(df.loc[i, 'spouse_id'], downcast='integer', errors='coerce')
if who in df['id'].values:
g.add_edge(pdt.Edge(pdt.Node(f"{df.loc[df['id']==who, 'id'].values[0]}"), nod, dir='none',color=col_w)) #label='m',
for i in range(df.shape[0]):
cc=col_f
if df.loc[i, 'gender']==1: cc=col_m
sstr=f"{df.loc[i, 'name']}"
if notnull(df.loc[i, 'surname']): sstr=f"{sstr} {df.loc[i, 'surname']}"
if notnull(df.loc[i, 'date_birth']):
if isinstance(df.loc[i, 'date_birth'], int) or isinstance(df.loc[i, 'date_birth'], str):
sstr=f"{sstr}\nB: {df.loc[i, 'date_birth']}"
else:
sstr=f"{sstr}\nB: {df.loc[i, 'date_birth'].strftime('%d %b %Y') }" # %H:%M:%S
d=df.loc[i, 'date_death']
# import numpy as np
if notnull(d):
# print("ddd" , isna(d), isnull(d), d, type(d) )
sstr=f"{sstr}\nD: {d}"
if notnull(df.loc[i, 'place_birth']):
sstr=f"{sstr}\nPoB: {df.loc[i, 'place_birth']} ({df.loc[i, 'country_birth']})"
if not (isnull(df.loc[i, 'comments']) or df.loc[i, 'comments']==''):
sstr=f"{sstr}\nComments: {df.loc[i, 'comments']}"
nod=pdt.Node(f"{df.loc[i, 'id']}", label=sstr, shape='box', color=cc, style='rounded')
g.add_node(nod)
import io
buf = io.BytesIO()
g.write_png(buf)
# plt.savefig(buf, format='png')
buf.seek(0)
img_g = Image_PIL.open(buf)
img = ImageTk.PhotoImage(img_g, master=frame)
canvas.create_image(0, 0, image=img)
#scrollconfig
canvas.config(scrollregion=canvas.bbox(tk.ALL))
xscrollbar.config(command=canvas.xview)
yscrollbar.config(command=canvas.yview)
frame.grid(row=0, column=0)
frame.grid_rowconfigure(0, weight=1)
frame.grid_columnconfigure(0, weight=1)
#menu
menubar = Menu(top)
top.config(menu=menubar)
menu = Menu(menubar)
def file_save():
from tkinter.filedialog import asksaveasfilename
f = asksaveasfilename(title = "Select file",filetypes = (("png files","*.png"),("all files","*.*")), defaultextension='.png')
if f is None: # asksaveasfile return `None` if dialog closed with "cancel".
return
img_g.save(f)
menu.add_command(label='Save',command=file_save)
menu.add_command(label='Close',command=top.destroy)
# menu.add_command(label="Exit", command=window.quit)
menubar.add_cascade(label="File", menu=menu)
top.mainloop()
def search_command():
for i in list1.tree.get_children():
list1.tree.delete(i)
# list1._build_tree(c_header, pddf(columns=c_header))
c_list=pddf(backend.search(name_text.get(),surname_text.get(),year_text.get(),gen_text.get()), columns=c_header)
list1._build_tree(c_header, c_list)
'name',
'surname',
'date_birth',
'date_death',
'mother_id',
'father_id',
'spouse_id',
'generation',
'cluster',
'gender',
'place_birth',
'country_birth',
'country',
'comments'
]
c_list=pddf(backend.view(), columns=c_header)
list1 = MultiColumnListbox(c_header, c_list)
''' fifth row '''
fr5 = Frame(window)
fr5.pack(fill=X, side=TOP)
# '''Buttons'''
b_clear=HoverButton(fr5,text="Clear Fields", width=12,command=clear_command, activebackground='SlateGray3')
b_clear.pack(side=tk.LEFT, pady=15, padx=10)
b3=HoverButton(fr5,text="Add entry", width=12,command=new_command, activebackground='SlateGray3')
b3.pack(side=tk.LEFT, pady=15, padx=10)
def draw_tree():
from PIL import ImageTk, Image as Image_PIL
c_header=['id',
'name',
'surname',
'date_birth',
'date_death',
'mother_id',
'father_id',
'spouse_id',
'generation',
'cluster',
'gender',
'place_birth',
'country_birth',
'country',
'comments'
]
# Image
import pydotplus as pdt
import time
# import pandas as pd
df=pddf(backend.view(), columns=c_header)
for c in ['id',
'mother_id',
'father_id',
'spouse_id',
'generation',
'cluster',
'gender' ]:
df[c]=to_numeric(df[c], downcast='integer',errors='coerce')
# df[c]=df[c].astype(int)
# df=pd.read_excel('c:/a_tree/gen_tree.xlsx', sheet_name=nn)
# print(df)
min_level=int(df['generation'].min())
max_level=int(df['generation'].max())
g = pdt.Dot(graph_type='digraph', compound='true', rankdir='TB',newrank = 'true')
i=0
edges = []
nodes=[]
# create all edges
#edges = [(1,2), (1,3), (2,4), (2,5), (3,5)]
#nodes = [(1, "A", "r"), (2, "B", "g"), (3, "C", "g"), (4, "D", "r"), (5, "E", "g")]
#for e in edges:
# g.add_edge(pdt.Edge(e[0], e[1], color=e[2]))
#
#for n in nodes:
# node = pdt.Node(name=n[0], label= n[1], fillcolor=n[2], style="filled" )
# g.add_node(node)
#
#
from ast import literal_eval as make_tuple
nl=max_level-min_level+1
s=[None]*(max_level+1)
for i in range(min_level,max_level+1):
s = pdt.Subgraph(i,rank='same')
for j in df.loc[df['generation']==i, 'id'].values :
# s.add_node(pdt.Node(f"{df.loc[df['id']==j, 'Name'].values[0]} {df.loc[df['id']==j, 'Surname'].values[0]}"))
s.add_node(pdt.Node(f"{df.loc[df['id']==j, 'id'].values[0]}"))
g.add_subgraph(s)
#min_cl=df['Cluster2'].min()
#max_cl=df['Cluster2'].max()
#nl=max_cl-min_cl+1
#ss=[None]*(max_cl+1)
#for i in range(min_cl,max_cl+1):
# if i>0:
# ss = pdt.Cluster(f'c2_{i}', color="blue",rankdir="TB")
# for j in df.loc[df['Cluster2']==i, 'id'].values :
# # s.add_node(pdt.Node(f"{df.loc[df['id']==j, 'Name'].values[0]} {df.loc[df['id']==j, 'Surname'].values[0]}"))
# ss.add_node(pdt.Node(f"{df.loc[df['id']==j, 'id'].values[0]}"))
# g.add_subgraph(ss)
min_cl=int(df['cluster'].min())
max_cl=int(df['cluster'].max())
nl=max_cl-min_cl+1
ss=[None]*(max_cl+1)
for i in range(min_cl,max_cl+1):
ss = pdt.Cluster(f'c1_{i}', color="white",rankdir="TB")
for j in df.loc[df['cluster']==i, 'id'].values :
# s.add_node(pdt.Node(f"{df.loc[df['id']==j, 'Name'].values[0]} {df.loc[df['id']==j, 'Surname'].values[0]}"))
ss.add_node(pdt.Node(f"{df.loc[df['id']==j, 'id'].values[0]}"))
g.add_subgraph(ss)
for i in range(df.shape[0]):
m=df.loc[i, 'mother_id']
f=df.loc[i, 'father_id']
# if m in df['id'].values: edges.append((i,df.loc[df['id']==m, 'Name'].index[0],'orange'))
# if f in df['id'].values: edges.append((i,df.loc[df['id']==f, 'Name'].index[0],'blue'))
# nodes.append((i, df.loc[i, 'Name'], 'blue'))
# nod=pdt.Node(f"{df.loc[i, 'id']}",label=f"{df.loc[i, 'Name']} {df.loc[i, 'Surname']}", shape='record', color="blue", style='rounded')
nod=pdt.Node(f"{df.loc[i, 'id']}", shape='record', color="blue", style='rounded')
if m in df['id'].values: g.add_edge(pdt.Edge(pdt.Node(f"{df.loc[df['id']==m, 'id'].values[0]}"), nod, color='orange'))
if f in df['id'].values: g.add_edge(pdt.Edge(pdt.Node(f"{df.loc[df['id']==f, 'id'].values[0]}"), nod, color='#066dba'))
if notnull(df.loc[i, 'spouse_id']):
# rel=df.loc[i, 'relation to']
# # tuple(map(int, re.findall(r'[0-9]+', rel)))
# rel=rel.replace('(','').replace(')','').split(',')
# who=pd.to_numeric(rel[0]).astype(int)
who=to_numeric(df.loc[i, 'spouse_id'], downcast='integer', errors='coerce')
# how=rel[1]
# g.add_edge(pdt.Edge(pdt.Node(f"{df.loc[df['id']==who, 'Name'].values[0]} {df.loc[df['id']==who, 'Surname'].values[0]}"), nod, dir='none',color='#bf3fbf')) #label='m',
if who in df['id'].values:
# print('yes')
g.add_edge(pdt.Edge(pdt.Node(f"{df.loc[df['id']==who, 'id'].values[0]}"), nod, dir='none',color='#bf3fbf')) #label='m',
for i in range(df.shape[0]):
cc='orange'
if df.loc[i, 'gender']==1: cc='#066dba'
sstr=f"{df.loc[i, 'name']}"
if notnull(df.loc[i, 'surname']): sstr=f"{sstr} {df.loc[i, 'surname']}"
if notnull(df.loc[i, 'date_birth']):
if isinstance(df.loc[i, 'date_birth'], int) or isinstance(df.loc[i, 'date_birth'], str):
sstr=f"{sstr}\nB: {df.loc[i, 'date_birth']}"
else:
sstr=f"{sstr}\nB: {df.loc[i, 'date_birth'].strftime('%d %b %Y') }" # %H:%M:%S
d=df.loc[i, 'date_death']
# import numpy as np
if notnull(d):
# print("ddd" , isna(d), isnull(d), d, type(d) )
sstr=f"{sstr}\nD: {d}"
if notnull(df.loc[i, 'place_birth']):
sstr=f"{sstr}\nPoB: {df.loc[i, 'place_birth']} ({df.loc[i, 'country_birth']})"
if not (isnull(df.loc[i, 'comments']) or df.loc[i, 'comments']==''):
sstr=f"{sstr}\nComments: {df.loc[i, 'comments']}"
nod=pdt.Node(f"{df.loc[i, 'id']}", label=sstr, shape='box', color=cc, style='rounded')
# nod=pdt.Node(f"{df.loc[i, 'id']}", label=f"{df.loc[i, 'Name']} {df.loc[i, 'Surname']}\nB: {df.loc[i, 'DoB']}\nD: {df.loc[i, 'DoD']}\nPoB: {df.loc[i, 'Place of Birth']} ({df.loc[i, 'Country of Birth']})", shape='box', color=cc, style='rounded')
g.add_node(nod)
import io
buf = io.BytesIO()
g.write_png(buf)
# plt.savefig(buf, format='png')
buf.seek(0)
import base64
buf_img64 = base64.b64encode(buf.getvalue())
# buf_im_str= u'data:img/jpeg;base64,'+buf_img64.decode('utf-8')
buf_im_str=buf_img64.decode("utf-8")
# img_g = Image_PIL.open(buf)
# img=Img_IP(g.create_png())
# img=g.create_png()
# return send_file(buf, mimetype='image/PNG')
# print (buf_img64)
# return render_template('image.html' , img= buf_img64.decode('ascii'))
return render_template('image.html' , img= buf_im_str)
def poolDataInterpolate(dropMap,
df,
label,
channel,
path,
incCarte=5e-3,
nbReps=80,
startTime=2,
timeThresh=40,
threshOutlayer=17,
display=False):
for j, tmp in enumerate(label):
dataX = np.array([])
dataY = np.array([])
data = pddf()
print(tmp)
nbAdded = 0
p = df[0]
x0 = np.array(p.time / 3600, dtype=np.float64)
p = df[
len(df) -
4] #antepenultien run pour creer le range de temps sur lequel interpolle
xLast = np.array(p.time / 3600, dtype=np.float64)
#x2=np.arange(.5, min(max(x0)-min(x0),max(xLast)-min(xLast)), .5)
x2 = np.arange(.5, timeThresh, .5)
data['time'] = pd.Series(x2).values
for j, content in enumerate(dropMap[:, 1]):
y = []
x = []
if content == tmp:
if channel == 'RFP':
p = df[j]
y = np.array(
np.log(p.fluo_2_area * p.speed / p.size) -
np.log(p.fluo_2_area[0] * p.speed[0] / p.size[0]))
x = np.array(p.time / 3600)
x = x - x[0] #reset time
ystd = np.array(
np.divide(p.fluo_2_std + incCarte / 2,
p.fluo_2_median))
if channel == 'GFP':
p = df[j]
y = np.array(
np.log(p.fluo_3_area * p.speed / p.size) -
np.log(p.fluo_3_area[0] * p.speed[0] / p.size[0]))
x = np.array(p.time / 3600, dtype=np.float64)
x = x - x[0] #reset time
ystd = np.array(
np.divide(p.fluo_3_std + incCarte / 2,
p.fluo_3_median))
if channel == 'PVD':
p = df[j]
y = np.array(
np.log(p.fluo_1_area * p.speed / p.size) -
np.log(p.fluo_1_area[0] * p.speed[0] / p.size[0]))
x = np.array(p.time / 3600, dtype=np.float64)
x = x - x[0] #reset time
ystd = np.array(
np.divide(p.fluo_1_std + incCarte / 2,
p.fluo_1_median))
if channel == 'PVDoverGFP':
p = df[j]
y = np.array(p.fluo_1_area /
(p.fluo_3_area)) #no need of log here
x = np.array(p.time / 3600, dtype=np.float64)
x = x - x[0] #reset time
ystd = np.array(np.divide(p.fluo_1_std + incCarte / 2,
p.fluo_1_median),
dtype=np.float64)
#filtre outlayer
#idxStart=bisect.bisect(x, startTime)
#idxThres=bisect.bisect(x, min(max(x),timeThresh))
idxOutlayer = bisect.bisect(x, threshOutlayer)
if len(y) > idxOutlayer:
if y[idxOutlayer] > y[2] + 2 or threshOutlayer <= 0:
# measure at run 0 does not count
f = si.interp1d(x, y)
try:
y2 = f(x2)
data[str(j)] = pd.Series(y2).values
nbAdded += 1
except Exception as inst:
print('poolData() remove this data')
print(j)
print(type(inst)) # the exception instance
print(inst.args) # arguments stored in .args
print(inst)
print('x2')
print(x2)
print('y')
print(y)
print('x')
print(x)
#plot the growth curve with the fit and the derivative
if display == True:
pathgrowth = path + 'growth/'
if not os.path.exists(pathgrowth):
os.makedirs(pathgrowth)
#plt.plot(x2,y2,marker='o', linestyle='-',color='blue')
plt.plot(x, y, marker='o', linestyle='-', color='blue')
#ax1.set_ylim([np.log(low), np.log(high)])
#ax1.set_yticks(range(np.int(np.log(low)), np.int(np.log(high))))
plt.ylabel('log(fluo)')
plt.xlabel('time (h)')
plt.title(tmp + ' drp' + str(j) + ' area')
plt.savefig(pathgrowth + 'drp' + str(j) +
'_deriveAll_nbpt' + '_' + tmp,
format='pdf')
plt.show()
if nbAdded > nbReps:
break
data.to_csv(path + tmp + channel + 'Interp.csv', index=False)
def getDataHistogram(label, path, channel):
folder = path + 'resultIndiv/'
[dropMap, df, nn] = loadData(path)
stdgRate = pddf()
gRate = pddf()
lag = pddf()
stdlag = pddf()
yld = pddf()
stdyield = pddf()
for n, labelName in enumerate(label):
stdgRateList = []
gRateList = []
lagList = []
stdlagList = []
yieldList = []
stdyieldList = []
for file in os.listdir(folder):
if file.startswith(labelName + 'resultIndiv_'):
if file.endswith(channel + ".csv"):
try:
with open(folder + file, 'r') as f:
reader = csv.reader(f)
your_list = list(reader)
stdgRateList.append(
float(your_list[findIdx(your_list,
'max df std')][1]))
gRateList.append(
float(your_list[findIdx(your_list, 'max df')][1]))
lagList.append(
float(your_list[findIdx(your_list,
'lag time')][1]))
stdlagList.append(
float(your_list[findIdx(your_list,
'lag time std')][1]))
yieldList.append(
float(your_list[findIdx(your_list, 'max y')][1]))
stdyieldList.append(
float(your_list[findIdx(your_list,
'max y std')][1]))
except Exception as inst:
donothing = 0
print(labelName)
print(folder + file)
print(type(inst)) # the exception instance
df = pddf({labelName: gRateList})
gRate = pd.concat([gRate, df], axis=1)
df = pddf({labelName: lagList})
lag = pd.concat([lag, df], axis=1)
df = pddf({labelName: yieldList})
yld = pd.concat([yld, df], axis=1)
df = pddf({labelName: stdgRateList})
stdgRate = pd.concat([stdgRate, df], axis=1)
df = pddf({labelName: stdlagList})
stdlag = pd.concat([stdlag, df], axis=1)
df = pddf({labelName: stdyieldList})
stdyield = pd.concat([stdyield, df], axis=1)
return [stdgRate, gRate, lag, stdlag, yld, stdyield]
def findIdx(your_list, item):
lst = pddf(your_list)
lst2 = lst[0].values
lst3 = lst2.tolist()
return lst3.index(item)