如何在新的 Tkinter window 中显示导入的数据框？
How do i display imported dataframe in new Tkinter window?
所以，我目前正在创建一个程序来对基本 excel 数据进行机器学习。但是我 运行 遇到了导入数据框后如何显示它的问题。
基本上，我需要在 page/window 上有一个名为“Dataframe”的“PageOutlier”按钮，它会提示一个新的 window，它显示我当前的数据框并使用最新信息进行更新来自“open_file()”函数。
我 运行 遇到的问题是：
如何在这种初始化框架的方式中创建一个新的（弹出）window？
如何在这个新 window 中最好地显示可能的 loaded/read 数据帧？
如何确保这个新 window 中显示的数据框是来自 运行 函数“open_file()”的最新数据框？
我试过搜索 Whosebug 和其他各种资源，但似乎找不到显示最新数据框的方法，所以我希望有人对此有解决方案。
虚拟数据可以是所有 excel 具有适合“open_file()”函数的数值的文件。
import tkinter as tk
from tkinter import ttk
from tkinter import filedialog
from tkinter.filedialog import askopenfilename
from tkinter.messagebox import showinfo, showwarning, askquestion
from tkinter import OptionMenu
from tkinter import StringVar
from matplotlib.backends.backend_tkagg import FigureCanvasTkAgg, NavigationToolbar2Tk
from matplotlib.figure import Figure
import matplotlib.pyplot as plt
from sklearn.metrics import silhouette_score
from sklearn.cluster import KMeans
import sklearn.cluster as cluster
import scipy.spatial.distance as sdist
from sklearn.ensemble import IsolationForest
import pandas as pd
import numpy as np
import seaborn as sn
from sklearn.preprocessing import MinMaxScaler
from sklearn.preprocessing import StandardScaler
RANDOM_STATE = 42 #used to help randomly select the data points
LARGE_FONT = ("Verdana", 12)
low_memory=False
df = None
df_nan = None
df_r = None
check_transformer = None
df_output = None
#f = Figure(figsize=(5,5), dpi=100)
#a = f.add_subplot(111)
def Credits():
showinfo("INFO",
'''
━━━━━━•❍•°•❍•°•❍•━━━━━━
Goodbye & have a nice day!
━━━━━•❍•°•❍•°•❍•━━━━━━
''')
app.destroy()
def RunUC():
showwarning("Warning", "Under construction")
def open_file():
global df, df_nan, df_r, df_output, best_n_clusters
df = None
df_nan = None
df_r = None
check_transformer = None
df_output = None
ClusterNum = None
C_Number = None
name = askopenfilename()
if name:
df = pd.read_excel(name, sep=';')
showinfo("INFO", "DataFrame created")
def Resetter():
global df, df_nan, df_r, df_output, ClusterNum, check_transformer
df = None
df_nan = None
df_r = None
check_transformer = None
df_output = None
ClusterNum = None
showinfo("INFO", "Dataframe reset")
def NaNifier():
global df, df_nan
df_nan = None
if df is None:
showwarning("Warning", "Read file first")
else:
df_nan = df
df_nan.replace(np.nan, 0, inplace=True)
showinfo("INFO", "Blank values replaced by zero values")
def data_remover():
global df_nan, df_r
if df is None:
#print("Read file first")
showwarning("Warning", "Read file first")
elif df_nan is None:
#print("Read file first")
showwarning("Warning", "Replace blank values first")
elif df_r is not None:
showwarning("Warning", "Data already cleaned")
else:
for y in df_nan.columns:
if df_nan[y].dtype != np.float64:
df_r = df_nan.select_dtypes(include=['float64', 'int64'])
elif df_nan[y].dtype != np.int64:
df_r = df_nan.select_dtypes(include=['float64', 'int64'])
else:
df_r = df_nan
showinfo("INFO", "Data cleaned")
def MM_data_transformer():
global df, df_nan, df_r, check_transformer
MM_scaler = MinMaxScaler()
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("Warning", "Clean data first")
elif check_transformer is not None:
showwarning("Warning", "Data already transformed")
else:
check_transformer = 1337 # Check to see if data is transformed
df_r = df_r.astype(float)
df_r = pd.DataFrame(MM_scaler.fit_transform(df_r),columns = df_r.columns) #Scaling data
df_r = df_r.round(2) #Rounding data
print(df_r)
showinfo("INFO", "Data scaled")
def S_data_transformer():
global df, df_nan, df_r, check_transformer
S_scaler = StandardScaler()
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("Warning", "Clean data first")
elif check_transformer is not None:
showwarning("Warning", "Data already transformed")
else:
check_transformer = 1337 # Check to see if data is transformed
df_r = df_r.astype(float)
df_r = pd.DataFrame(S_scaler.fit_transform(df_r),columns = df_r.columns) #Scaling data
df_r = df_r.round(2) #Rounding data
print(df_r)
showinfo("INFO", "Data scaled")
def RunSilHou():
#Silhouette Score for further clustering
global df, df_nan, df_r, df_output, ClusterNum
ClusterNum = None
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif df_output is not None:
showwarning("WARNING", "Already computed KMeans")
else:
n_samples, n_features = df_r.shape
procent_samples = n_samples
#procent_samples = n_samples*0.10 #10% of the dataset for sample_size
procent_samples = int(procent_samples)
sil_score_max = -1 #this is the minimum possible score
for n_clusters in range(2,10):
model = KMeans(n_clusters = n_clusters, n_jobs = 8, init='k-means++', max_iter=100, n_init=1)
labels = model.fit_predict(df_r)
sil_score = silhouette_score(df_r, labels, sample_size = procent_samples, random_state = RANDOM_STATE)
print("The average silhouette score for %i clusters is %0.10f" %(n_clusters, sil_score))
if sil_score > sil_score_max:
sil_score_max = sil_score
ClusterNum = n_clusters
Cluster_text = "Silhouette Score calculated %i clusters" %(ClusterNum)
showinfo(title="Results", message=Cluster_text)
def RunKM_clustering():
global df, df_nan, df_r, df_output, ClusterNum
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif df_output is not None:
showwarning("WARNING", "Already computed KMeans")
else:
df_output = df
kmeans = cluster.KMeans(n_jobs = 8, n_clusters = ClusterNum, init = 'k-means++', random_state = RANDOM_STATE).fit(df_r)
centroids = kmeans.cluster_centers_
dists = pd.DataFrame(
sdist.cdist(df_r, centroids),
columns=['dist_{}'.format(i) for i in range(len(centroids))],
index=df_output.index)
df_output = pd.concat([df_r, dists], axis=1)
df_output['cluster_number'] = kmeans.labels_
df_output_columns = df_output.filter(regex=('dist_')).columns
df_output['dists'] = df_output[df_output_columns].min(axis=1)
df_output = df_output[df_output.columns.drop(list(df_output.filter(regex='dist_')))]
showinfo("INFO", "KMeans clusters Calculated")
def RunKM_outlier():
global df, df_nan, df_r, df_output
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif df_output is not None:
kmeans_one = cluster.KMeans(n_jobs = 8, n_clusters = 1, init = 'k-means++', random_state = RANDOM_STATE).fit(df_r)
centroids = kmeans_one.cluster_centers_
dist = pd.DataFrame(
sdist.cdist(df_r, centroids),
columns=['dist_clust_outlier'.format(i) for i in range(len(centroids))],
index=df_output.index)
df_output = pd.concat([df_output, dist], axis=1)
else:
df_output = df
kmeans_one = cluster.KMeans(n_jobs = 8, n_clusters = 1, init = 'k-means++', random_state = RANDOM_STATE).fit(df_r)
centroids = kmeans_one.cluster_centers_
dist = pd.DataFrame(
sdist.cdist(df_r, centroids),
columns=['dist_clust_outlier'.format(i) for i in range(len(centroids))],
index=df_output.index)
df_output = pd.concat([df_output, dist], axis=1)
def RunIsoForest():
global df, df_nan, df_r, df_output, ContNum
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif ContNum is None:
showwarning('Warning', "Define Contamination Value")
elif df_output is not None:
iso_columns = df_r.columns
clf = IsolationForest(max_samples="auto", n_jobs=-1, random_state=RANDOM_STATE, behaviour="new", contamination=ContNum)
clf.fit(df_r[iso_columns])
iso_pred = clf.predict(df_r[iso_columns])
df_output['isolation_class'] = iso_pred
showinfo("INFO", "IsolationForest Calculated")
else:
df_output = df
iso_columns = df_r.columns
clf = IsolationForest(max_samples="auto", n_jobs=-1, random_state=RANDOM_STATE, behaviour="new", contamination=ContNum)
clf.fit(df_r[iso_columns])
iso_pred = clf.predict(df_r[iso_columns])
df_output['isolation_class'] = iso_pred
showinfo("INFO", "IsolationForest Calculated")
def save_in_new_file():
global df, df_nan, df_r, df_output
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif df_output is None:
showwarning("WARNING", "Compute file first")
else:
df_output.to_csv("output.csv", sep=';', index=0, mode='w')
showinfo("INFO", "DataFrame saved")
def ClusterDefinition():
global ClusterNum
ClusterNum = Clus_Number.get()
ClusterNum = int(ClusterNum)
print(ClusterNum)
Cluster_text = "KMeans clustering set to %i clusters" %(ClusterNum)
showinfo(title="Results", message=Cluster_text)
def ContaminationDefinition():
global ContNum
ContNum = Cont_Number.get()
ContNum = int(ContNum)
ContNum = ContNum/100 # to return integer to 0.0X state
print(ContNum)
Contamination_text = "Contamination value set to %0.2f" %(ContNum)
showinfo(title="Results", message=Contamination_text)
class Mainapp(tk.Tk):
def __init__(self, *args, **kwargs):
tk.Tk.__init__(self, *args, **kwargs)
#tk.Tk.iconbitmap(self, default="iconimage_kmeans.ico") #Icon for program
tk.Tk.wm_title(self, "Advanched analytics")
container = tk.Frame(self)
container.pack(side="top", fill="both", expand = True)
container.grid_rowconfigure(0, weight=1)
container.grid_columnconfigure(0, weight=1)
menubar = tk.Menu(container)
filemenu_file = tk.Menu(menubar, tearoff=0)
filemenu_file.add_command(label="Open", command=open_file)
filemenu_file.add_command(label="Save", command=save_in_new_file)
filemenu_file.add_separator()
filemenu_file.add_command(label="Exit", command=Credits)
menubar.add_cascade(label="File", menu=filemenu_file)
filemenu_edit = tk.Menu(menubar, tearoff=0)
submenu_t = tk.Menu(container, tearoff=0)
submenu_s = tk.Menu(container, tearoff=0)
submenu_t.add_command(label="Blank values", command=NaNifier)
submenu_t.add_command(label="Non-numeric values", command=data_remover)
submenu_s.add_command(label="MinMax scaler", command=MM_data_transformer)
submenu_s.add_command(label="Standard scaler", command=S_data_transformer)
filemenu_edit.add_cascade(label='Data transformating', menu=submenu_t, underline=0)
filemenu_edit.add_cascade(label='Data scaling', menu=submenu_s, underline=0)
filemenu_edit.add_separator()
filemenu_edit.add_command(label="Clear dataframe", command=Resetter)
menubar.add_cascade(label="Edit", menu=filemenu_edit)
filemenu_help = tk.Menu(menubar, tearoff=0)
filemenu_help.add_command(label="Function descriptions", command=RunUC) # Make help txt. file
filemenu_help.add_command(label="Data cleaning steps", command=RunUC)
menubar.add_cascade(label="Help", menu=filemenu_help)
tk.Tk.config(self, menu=menubar)
self.frames = {}
for F in (StartPage, PageCluster, PageOutlier, PageElbow, DataPage):
frame = F(container, self)
self.frames[F] = frame
frame.grid(row=0, column=0, sticky="nsew")
self.show_frame(StartPage)
def show_frame(self, cont):
frame = self.frames[cont]
frame.tkraise()
class StartPage(tk.Frame):
def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
label = tk.Label(self, text=
"Advanched analytics", font=LARGE_FONT)
label.pack(pady=10, padx=10)
button1 = ttk.Button(self, text="Clustering",
command=lambda: controller.show_frame(PageCluster))
button1.pack(fill='x')
button2 = ttk.Button(self, text="Outlier Detection",
command=lambda: controller.show_frame(PageOutlier))
button2.pack(fill='x')
class PageCluster(tk.Frame):
def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
label = tk.Label(self, text="Clustering", font=LARGE_FONT)
label.pack(pady=10, padx=10)
button1 = ttk.Button(self, text='Run Silhouette Score Calculation', command=RunSilHou)
button1.pack(fill='x')
button2 = ttk.Button(self, text="Elbow Method Calculation",
command=lambda: controller.show_frame(PageElbow))
button2.pack(fill='x')
button3 = ttk.Button(self, text='Run KMeans', command=RunKM_clustering)
button3.pack(fill='x')
button4 = ttk.Button(self, text="Back",
command=lambda: controller.show_frame(StartPage))
button4.pack(fill='x')
class PageElbow(tk.Frame):
def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
label = tk.Label(self, text="Elbow Method", font=LARGE_FONT)
label.pack(pady=10, padx=10)
global ClusterNum, C_Number
def RunElbowMethod():
f = None
f = plt.figure(figsize=(10, 8))
wcss = []
for i in range(1, 11):
kmeans = KMeans(n_clusters = i, init = 'k-means++', random_state = RANDOM_STATE)
kmeans.fit(df)
wcss.append(kmeans.inertia_)
plt.plot(range(1, 11), wcss)
plt.title('The Elbow Method')
plt.xlabel('Number of clusters')
plt.ylabel('WCSS')
plt.draw()
canvas = FigureCanvasTkAgg(f, self)
canvas.draw()
canvas.get_tk_widget().pack(side=tk.BOTTOM, fill=tk.BOTH, expand=True)
toolbar = NavigationToolbar2Tk(canvas, self)
toolbar.update()
canvas._tkcanvas.pack(side=tk.TOP, fill=tk.BOTH, expand=True)
ClusterOptions = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
Clus_Number = StringVar(self)
Clus_Number.set(ClusterOptions[0]) # default value
button1 = ttk.Button(self, text="Run Elbow Method", command=RunElbowMethod)
button1.pack(fill='x')
dropdown1 = tk.OptionMenu(self, Clus_Number, *ClusterOptions)
dropdown1.pack()
button2 = ttk.Button(self, text="Define Number of Clusters", command=ClusterDefinition)
button2.pack(fill='x')
button3 = ttk.Button(self, text="Back",
command=lambda: controller.show_frame(PageCluster))
button3.pack(fill='x')
class PageOutlier(tk.Frame):
def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
label = tk.Label(self, text="Outlier Detection", font=LARGE_FONT)
label.pack(pady=10, padx=10)
global ContNum, Cont_Number
ContOptions = np.arange(1, 100, 1)
Cont_Number = StringVar(self)
Cont_Number.set(ContOptions[0]) # default value
dropdown1 = tk.OptionMenu(self, Cont_Number, *ContOptions)
dropdown1.pack()
button2 = ttk.Button(self, text="Define Contamination Value (0.X Decimals)", command=ContaminationDefinition)
button2.pack(fill='x')
button3 = ttk.Button(self, text='Run IsolationForest Outlier Detection', command=RunIsoForest)
button3.pack(fill='x')
button4 = ttk.Button(self, text='Dataframe', command=RunUC)
button4.pack(fill='x')
button5 = ttk.Button(self, text="Back",
command=lambda: controller.show_frame(StartPage))
button5.pack(fill='x')
app = Mainapp()
app.geometry('500x400')
app.mainloop()
#app.destroy()
我在下面创建了一个演示，其中包含有关如何执行此操作的评论。在下面的代码中，我创建了一个名为 ExcelWindow 的 class，它继承自 tk.Toplevel，这将打开一个新的 window。我还在这个 class 中传递了一个 DataFrame 参数。一旦创建了 ExcelWindow 中的 Treeview Widget，它就会将您的数据框插入其中。因此，您可以在新的 window 中查看数据框。您也可以使用不同的文件打开任意多个 windows！
要演示下面的代码，请在您的 PC 上使用以下数据创建一个 excel 文件 (.xlsx)：
然后运行下面的代码并点击浏览按钮。在 windows 资源管理器中单击文件后，单击“加载文件”按钮，然后会出现一个包含 excel 数据的新 Tkinter window。如前所述，所有注释都在代码中。
import tkinter as tk
from tkinter import filedialog, messagebox, ttk
import pandas as pd
# a video tutorial for a similar question can be found here https://www.youtube.com/watch?v=PgLjwl6Br0k
# initalise the tkinter GUI
root = tk.Tk()
root.geometry("500x500") # set the root dimensions
root.pack_propagate(False) # tells the root to not let the widgets inside it determine its size.
root.resizable(0, 0) # makes the root window fixed in size.
# Frame for open file dialog
file_frame = tk.LabelFrame(root, text="Open File")
file_frame.place(height=100, width=400, rely=0.65, relx=0)
# Buttons
button1 = tk.Button(file_frame, text="Browse A File", command=lambda: File_dialog())
button1.place(rely=0.65, relx=0.50)
button2 = tk.Button(file_frame, text="Load File", command=lambda: Load_excel_data())
button2.place(rely=0.65, relx=0.30)
# The file/file path text
label_file = ttk.Label(file_frame, text="No File Selected")
label_file.place(rely=0, relx=0)
# this will open the windows explorer so you can select your file
def File_dialog():
filename = filedialog.askopenfilename(initialdir="/",
title="Select A File",
filetype=(("xlsx files", "*.xlsx"),("All Files", "*.*")))
label_file["text"] = filename
return None
# this will load your excel file into the excel window
def Load_excel_data():
file_path = label_file["text"] # get the text from the path you provided
try:
excel_filename = r"{}".format(file_path)
if excel_filename[-4:] == ".csv": # if its a csv file use read_csv
df = pd.read_csv(excel_filename)
else: # otherwise use read_excel
df = pd.read_excel(excel_filename)
except ValueError:
tk.messagebox.showerror("Information", "The file you have chosen is invalid")
return None
except FileNotFoundError:
tk.messagebox.showerror("Information", f"No such file as {file_path}")
return None
# pass the dataframe into the class ExcelWindow
ExcelWindow(df)
class ExcelWindow(tk.Toplevel): # Inherits from tk.Toplevel
def __init__(self, your_dataframe): # the dataframe you passed through is here
super().__init__()
# Frame for TreeView
frame1 = tk.LabelFrame(self, text="Excel Data")
frame1.pack(fill="both", expand="true")
frame1.pack_propagate(0)
# the size of the window.
self.geometry("500x500")
self.resizable(0, 0)
self.title("Your Excel File in Tkinter") # the window title
# This creates your Treeview widget.
tv1 = ttk.Treeview(frame1)
tv1.place(relheight=1, relwidth=1) # set the height and width of the widget to 100% of its container (frame1).
treescrolly = tk.Scrollbar(frame1, orient="vertical", command=tv1.yview) # command means update the yaxis view of the widget
treescrollx = tk.Scrollbar(frame1, orient="horizontal", command=tv1.xview) # command means update the xaxis view of the widget
tv1.configure(xscrollcommand=treescrollx.set, yscrollcommand=treescrolly.set) # assign the scrollbars to the Treeview Widget
treescrollx.pack(side="bottom", fill="x") # make the scrollbar fill the x axis of the Treeview widget
treescrolly.pack(side="right", fill="y") # make the scrollbar fill the y axis of the Treeview widget
# this loads the dataframe into the treeview widget
tv1["column"] = list(your_dataframe.columns)
tv1["show"] = "headings"
for column in tv1["columns"]:
tv1.heading(column, text=column) # let the column heading = column name
df_rows = your_dataframe.to_numpy().tolist() # turns the dataframe into a list of lists
for row in df_rows:
tv1.insert("", "end", values=row) # inserts each list into the treeview. For parameters see https://docs.python.org/3/library/tkinter.ttk.html#tkinter.ttk.Treeview.insert
root.mainloop()
我通过结合来自 Whosebug 的一些答案自己弄明白了。
基本上通过一个函数而不是我的“普通”框架创建一个新的 window，然后使用 command=DisplayDF 创建一个按钮。这确保它只看到最新的数据帧，并且在调用时不会因为在导入之前被引用而产生错误。
def DisplayDF():
global df
# create child window
win = Toplevel()
# display message
message = "Current output (Max 100 rows)"
tk.Label(win, text=message).pack()
text = tk.Text(win)
text.insert(tk.END, str(df.head(100)))
text.pack()
所以，我目前正在创建一个程序来对基本 excel 数据进行机器学习。但是我 运行 遇到了导入数据框后如何显示它的问题。
基本上，我需要在 page/window 上有一个名为“Dataframe”的“PageOutlier”按钮，它会提示一个新的 window，它显示我当前的数据框并使用最新信息进行更新来自“open_file()”函数。
我 运行 遇到的问题是：
如何在这种初始化框架的方式中创建一个新的（弹出）window？
如何在这个新 window 中最好地显示可能的 loaded/read 数据帧？
如何确保这个新 window 中显示的数据框是来自 运行 函数“open_file()”的最新数据框？
我试过搜索 Whosebug 和其他各种资源，但似乎找不到显示最新数据框的方法，所以我希望有人对此有解决方案。
虚拟数据可以是所有 excel 具有适合“open_file()”函数的数值的文件。
import tkinter as tk
from tkinter import ttk
from tkinter import filedialog
from tkinter.filedialog import askopenfilename
from tkinter.messagebox import showinfo, showwarning, askquestion
from tkinter import OptionMenu
from tkinter import StringVar
from matplotlib.backends.backend_tkagg import FigureCanvasTkAgg, NavigationToolbar2Tk
from matplotlib.figure import Figure
import matplotlib.pyplot as plt
from sklearn.metrics import silhouette_score
from sklearn.cluster import KMeans
import sklearn.cluster as cluster
import scipy.spatial.distance as sdist
from sklearn.ensemble import IsolationForest
import pandas as pd
import numpy as np
import seaborn as sn
from sklearn.preprocessing import MinMaxScaler
from sklearn.preprocessing import StandardScaler
RANDOM_STATE = 42 #used to help randomly select the data points
LARGE_FONT = ("Verdana", 12)
low_memory=False
df = None
df_nan = None
df_r = None
check_transformer = None
df_output = None
#f = Figure(figsize=(5,5), dpi=100)
#a = f.add_subplot(111)
def Credits():
showinfo("INFO",
'''
━━━━━━•❍•°•❍•°•❍•━━━━━━
Goodbye & have a nice day!
━━━━━•❍•°•❍•°•❍•━━━━━━
''')
app.destroy()
def RunUC():
showwarning("Warning", "Under construction")
def open_file():
global df, df_nan, df_r, df_output, best_n_clusters
df = None
df_nan = None
df_r = None
check_transformer = None
df_output = None
ClusterNum = None
C_Number = None
name = askopenfilename()
if name:
df = pd.read_excel(name, sep=';')
showinfo("INFO", "DataFrame created")
def Resetter():
global df, df_nan, df_r, df_output, ClusterNum, check_transformer
df = None
df_nan = None
df_r = None
check_transformer = None
df_output = None
ClusterNum = None
showinfo("INFO", "Dataframe reset")
def NaNifier():
global df, df_nan
df_nan = None
if df is None:
showwarning("Warning", "Read file first")
else:
df_nan = df
df_nan.replace(np.nan, 0, inplace=True)
showinfo("INFO", "Blank values replaced by zero values")
def data_remover():
global df_nan, df_r
if df is None:
#print("Read file first")
showwarning("Warning", "Read file first")
elif df_nan is None:
#print("Read file first")
showwarning("Warning", "Replace blank values first")
elif df_r is not None:
showwarning("Warning", "Data already cleaned")
else:
for y in df_nan.columns:
if df_nan[y].dtype != np.float64:
df_r = df_nan.select_dtypes(include=['float64', 'int64'])
elif df_nan[y].dtype != np.int64:
df_r = df_nan.select_dtypes(include=['float64', 'int64'])
else:
df_r = df_nan
showinfo("INFO", "Data cleaned")
def MM_data_transformer():
global df, df_nan, df_r, check_transformer
MM_scaler = MinMaxScaler()
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("Warning", "Clean data first")
elif check_transformer is not None:
showwarning("Warning", "Data already transformed")
else:
check_transformer = 1337 # Check to see if data is transformed
df_r = df_r.astype(float)
df_r = pd.DataFrame(MM_scaler.fit_transform(df_r),columns = df_r.columns) #Scaling data
df_r = df_r.round(2) #Rounding data
print(df_r)
showinfo("INFO", "Data scaled")
def S_data_transformer():
global df, df_nan, df_r, check_transformer
S_scaler = StandardScaler()
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("Warning", "Clean data first")
elif check_transformer is not None:
showwarning("Warning", "Data already transformed")
else:
check_transformer = 1337 # Check to see if data is transformed
df_r = df_r.astype(float)
df_r = pd.DataFrame(S_scaler.fit_transform(df_r),columns = df_r.columns) #Scaling data
df_r = df_r.round(2) #Rounding data
print(df_r)
showinfo("INFO", "Data scaled")
def RunSilHou():
#Silhouette Score for further clustering
global df, df_nan, df_r, df_output, ClusterNum
ClusterNum = None
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif df_output is not None:
showwarning("WARNING", "Already computed KMeans")
else:
n_samples, n_features = df_r.shape
procent_samples = n_samples
#procent_samples = n_samples*0.10 #10% of the dataset for sample_size
procent_samples = int(procent_samples)
sil_score_max = -1 #this is the minimum possible score
for n_clusters in range(2,10):
model = KMeans(n_clusters = n_clusters, n_jobs = 8, init='k-means++', max_iter=100, n_init=1)
labels = model.fit_predict(df_r)
sil_score = silhouette_score(df_r, labels, sample_size = procent_samples, random_state = RANDOM_STATE)
print("The average silhouette score for %i clusters is %0.10f" %(n_clusters, sil_score))
if sil_score > sil_score_max:
sil_score_max = sil_score
ClusterNum = n_clusters
Cluster_text = "Silhouette Score calculated %i clusters" %(ClusterNum)
showinfo(title="Results", message=Cluster_text)
def RunKM_clustering():
global df, df_nan, df_r, df_output, ClusterNum
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif df_output is not None:
showwarning("WARNING", "Already computed KMeans")
else:
df_output = df
kmeans = cluster.KMeans(n_jobs = 8, n_clusters = ClusterNum, init = 'k-means++', random_state = RANDOM_STATE).fit(df_r)
centroids = kmeans.cluster_centers_
dists = pd.DataFrame(
sdist.cdist(df_r, centroids),
columns=['dist_{}'.format(i) for i in range(len(centroids))],
index=df_output.index)
df_output = pd.concat([df_r, dists], axis=1)
df_output['cluster_number'] = kmeans.labels_
df_output_columns = df_output.filter(regex=('dist_')).columns
df_output['dists'] = df_output[df_output_columns].min(axis=1)
df_output = df_output[df_output.columns.drop(list(df_output.filter(regex='dist_')))]
showinfo("INFO", "KMeans clusters Calculated")
def RunKM_outlier():
global df, df_nan, df_r, df_output
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif df_output is not None:
kmeans_one = cluster.KMeans(n_jobs = 8, n_clusters = 1, init = 'k-means++', random_state = RANDOM_STATE).fit(df_r)
centroids = kmeans_one.cluster_centers_
dist = pd.DataFrame(
sdist.cdist(df_r, centroids),
columns=['dist_clust_outlier'.format(i) for i in range(len(centroids))],
index=df_output.index)
df_output = pd.concat([df_output, dist], axis=1)
else:
df_output = df
kmeans_one = cluster.KMeans(n_jobs = 8, n_clusters = 1, init = 'k-means++', random_state = RANDOM_STATE).fit(df_r)
centroids = kmeans_one.cluster_centers_
dist = pd.DataFrame(
sdist.cdist(df_r, centroids),
columns=['dist_clust_outlier'.format(i) for i in range(len(centroids))],
index=df_output.index)
df_output = pd.concat([df_output, dist], axis=1)
def RunIsoForest():
global df, df_nan, df_r, df_output, ContNum
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif ContNum is None:
showwarning('Warning', "Define Contamination Value")
elif df_output is not None:
iso_columns = df_r.columns
clf = IsolationForest(max_samples="auto", n_jobs=-1, random_state=RANDOM_STATE, behaviour="new", contamination=ContNum)
clf.fit(df_r[iso_columns])
iso_pred = clf.predict(df_r[iso_columns])
df_output['isolation_class'] = iso_pred
showinfo("INFO", "IsolationForest Calculated")
else:
df_output = df
iso_columns = df_r.columns
clf = IsolationForest(max_samples="auto", n_jobs=-1, random_state=RANDOM_STATE, behaviour="new", contamination=ContNum)
clf.fit(df_r[iso_columns])
iso_pred = clf.predict(df_r[iso_columns])
df_output['isolation_class'] = iso_pred
showinfo("INFO", "IsolationForest Calculated")
def save_in_new_file():
global df, df_nan, df_r, df_output
if df is None:
showwarning("Warning", "Read file first")
elif df_nan is None:
showwarning("Warning", "Replace blank values first")
elif df_r is None:
showwarning("WARNING", "Clean file first")
elif df_output is None:
showwarning("WARNING", "Compute file first")
else:
df_output.to_csv("output.csv", sep=';', index=0, mode='w')
showinfo("INFO", "DataFrame saved")
def ClusterDefinition():
global ClusterNum
ClusterNum = Clus_Number.get()
ClusterNum = int(ClusterNum)
print(ClusterNum)
Cluster_text = "KMeans clustering set to %i clusters" %(ClusterNum)
showinfo(title="Results", message=Cluster_text)
def ContaminationDefinition():
global ContNum
ContNum = Cont_Number.get()
ContNum = int(ContNum)
ContNum = ContNum/100 # to return integer to 0.0X state
print(ContNum)
Contamination_text = "Contamination value set to %0.2f" %(ContNum)
showinfo(title="Results", message=Contamination_text)
class Mainapp(tk.Tk):
def __init__(self, *args, **kwargs):
tk.Tk.__init__(self, *args, **kwargs)
#tk.Tk.iconbitmap(self, default="iconimage_kmeans.ico") #Icon for program
tk.Tk.wm_title(self, "Advanched analytics")
container = tk.Frame(self)
container.pack(side="top", fill="both", expand = True)
container.grid_rowconfigure(0, weight=1)
container.grid_columnconfigure(0, weight=1)
menubar = tk.Menu(container)
filemenu_file = tk.Menu(menubar, tearoff=0)
filemenu_file.add_command(label="Open", command=open_file)
filemenu_file.add_command(label="Save", command=save_in_new_file)
filemenu_file.add_separator()
filemenu_file.add_command(label="Exit", command=Credits)
menubar.add_cascade(label="File", menu=filemenu_file)
filemenu_edit = tk.Menu(menubar, tearoff=0)
submenu_t = tk.Menu(container, tearoff=0)
submenu_s = tk.Menu(container, tearoff=0)
submenu_t.add_command(label="Blank values", command=NaNifier)
submenu_t.add_command(label="Non-numeric values", command=data_remover)
submenu_s.add_command(label="MinMax scaler", command=MM_data_transformer)
submenu_s.add_command(label="Standard scaler", command=S_data_transformer)
filemenu_edit.add_cascade(label='Data transformating', menu=submenu_t, underline=0)
filemenu_edit.add_cascade(label='Data scaling', menu=submenu_s, underline=0)
filemenu_edit.add_separator()
filemenu_edit.add_command(label="Clear dataframe", command=Resetter)
menubar.add_cascade(label="Edit", menu=filemenu_edit)
filemenu_help = tk.Menu(menubar, tearoff=0)
filemenu_help.add_command(label="Function descriptions", command=RunUC) # Make help txt. file
filemenu_help.add_command(label="Data cleaning steps", command=RunUC)
menubar.add_cascade(label="Help", menu=filemenu_help)
tk.Tk.config(self, menu=menubar)
self.frames = {}
for F in (StartPage, PageCluster, PageOutlier, PageElbow, DataPage):
frame = F(container, self)
self.frames[F] = frame
frame.grid(row=0, column=0, sticky="nsew")
self.show_frame(StartPage)
def show_frame(self, cont):
frame = self.frames[cont]
frame.tkraise()
class StartPage(tk.Frame):
def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
label = tk.Label(self, text=
"Advanched analytics", font=LARGE_FONT)
label.pack(pady=10, padx=10)
button1 = ttk.Button(self, text="Clustering",
command=lambda: controller.show_frame(PageCluster))
button1.pack(fill='x')
button2 = ttk.Button(self, text="Outlier Detection",
command=lambda: controller.show_frame(PageOutlier))
button2.pack(fill='x')
class PageCluster(tk.Frame):
def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
label = tk.Label(self, text="Clustering", font=LARGE_FONT)
label.pack(pady=10, padx=10)
button1 = ttk.Button(self, text='Run Silhouette Score Calculation', command=RunSilHou)
button1.pack(fill='x')
button2 = ttk.Button(self, text="Elbow Method Calculation",
command=lambda: controller.show_frame(PageElbow))
button2.pack(fill='x')
button3 = ttk.Button(self, text='Run KMeans', command=RunKM_clustering)
button3.pack(fill='x')
button4 = ttk.Button(self, text="Back",
command=lambda: controller.show_frame(StartPage))
button4.pack(fill='x')
class PageElbow(tk.Frame):
def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
label = tk.Label(self, text="Elbow Method", font=LARGE_FONT)
label.pack(pady=10, padx=10)
global ClusterNum, C_Number
def RunElbowMethod():
f = None
f = plt.figure(figsize=(10, 8))
wcss = []
for i in range(1, 11):
kmeans = KMeans(n_clusters = i, init = 'k-means++', random_state = RANDOM_STATE)
kmeans.fit(df)
wcss.append(kmeans.inertia_)
plt.plot(range(1, 11), wcss)
plt.title('The Elbow Method')
plt.xlabel('Number of clusters')
plt.ylabel('WCSS')
plt.draw()
canvas = FigureCanvasTkAgg(f, self)
canvas.draw()
canvas.get_tk_widget().pack(side=tk.BOTTOM, fill=tk.BOTH, expand=True)
toolbar = NavigationToolbar2Tk(canvas, self)
toolbar.update()
canvas._tkcanvas.pack(side=tk.TOP, fill=tk.BOTH, expand=True)
ClusterOptions = [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
Clus_Number = StringVar(self)
Clus_Number.set(ClusterOptions[0]) # default value
button1 = ttk.Button(self, text="Run Elbow Method", command=RunElbowMethod)
button1.pack(fill='x')
dropdown1 = tk.OptionMenu(self, Clus_Number, *ClusterOptions)
dropdown1.pack()
button2 = ttk.Button(self, text="Define Number of Clusters", command=ClusterDefinition)
button2.pack(fill='x')
button3 = ttk.Button(self, text="Back",
command=lambda: controller.show_frame(PageCluster))
button3.pack(fill='x')
class PageOutlier(tk.Frame):
def __init__(self, parent, controller):
tk.Frame.__init__(self, parent)
label = tk.Label(self, text="Outlier Detection", font=LARGE_FONT)
label.pack(pady=10, padx=10)
global ContNum, Cont_Number
ContOptions = np.arange(1, 100, 1)
Cont_Number = StringVar(self)
Cont_Number.set(ContOptions[0]) # default value
dropdown1 = tk.OptionMenu(self, Cont_Number, *ContOptions)
dropdown1.pack()
button2 = ttk.Button(self, text="Define Contamination Value (0.X Decimals)", command=ContaminationDefinition)
button2.pack(fill='x')
button3 = ttk.Button(self, text='Run IsolationForest Outlier Detection', command=RunIsoForest)
button3.pack(fill='x')
button4 = ttk.Button(self, text='Dataframe', command=RunUC)
button4.pack(fill='x')
button5 = ttk.Button(self, text="Back",
command=lambda: controller.show_frame(StartPage))
button5.pack(fill='x')
app = Mainapp()
app.geometry('500x400')
app.mainloop()
#app.destroy()
我在下面创建了一个演示，其中包含有关如何执行此操作的评论。在下面的代码中，我创建了一个名为 ExcelWindow 的 class，它继承自 tk.Toplevel，这将打开一个新的 window。我还在这个 class 中传递了一个 DataFrame 参数。一旦创建了 ExcelWindow 中的 Treeview Widget，它就会将您的数据框插入其中。因此，您可以在新的 window 中查看数据框。您也可以使用不同的文件打开任意多个 windows！
要演示下面的代码，请在您的 PC 上使用以下数据创建一个 excel 文件 (.xlsx)：
然后运行下面的代码并点击浏览按钮。在 windows 资源管理器中单击文件后，单击“加载文件”按钮，然后会出现一个包含 excel 数据的新 Tkinter window。如前所述，所有注释都在代码中。
import tkinter as tk
from tkinter import filedialog, messagebox, ttk
import pandas as pd
# a video tutorial for a similar question can be found here https://www.youtube.com/watch?v=PgLjwl6Br0k
# initalise the tkinter GUI
root = tk.Tk()
root.geometry("500x500") # set the root dimensions
root.pack_propagate(False) # tells the root to not let the widgets inside it determine its size.
root.resizable(0, 0) # makes the root window fixed in size.
# Frame for open file dialog
file_frame = tk.LabelFrame(root, text="Open File")
file_frame.place(height=100, width=400, rely=0.65, relx=0)
# Buttons
button1 = tk.Button(file_frame, text="Browse A File", command=lambda: File_dialog())
button1.place(rely=0.65, relx=0.50)
button2 = tk.Button(file_frame, text="Load File", command=lambda: Load_excel_data())
button2.place(rely=0.65, relx=0.30)
# The file/file path text
label_file = ttk.Label(file_frame, text="No File Selected")
label_file.place(rely=0, relx=0)
# this will open the windows explorer so you can select your file
def File_dialog():
filename = filedialog.askopenfilename(initialdir="/",
title="Select A File",
filetype=(("xlsx files", "*.xlsx"),("All Files", "*.*")))
label_file["text"] = filename
return None
# this will load your excel file into the excel window
def Load_excel_data():
file_path = label_file["text"] # get the text from the path you provided
try:
excel_filename = r"{}".format(file_path)
if excel_filename[-4:] == ".csv": # if its a csv file use read_csv
df = pd.read_csv(excel_filename)
else: # otherwise use read_excel
df = pd.read_excel(excel_filename)
except ValueError:
tk.messagebox.showerror("Information", "The file you have chosen is invalid")
return None
except FileNotFoundError:
tk.messagebox.showerror("Information", f"No such file as {file_path}")
return None
# pass the dataframe into the class ExcelWindow
ExcelWindow(df)
class ExcelWindow(tk.Toplevel): # Inherits from tk.Toplevel
def __init__(self, your_dataframe): # the dataframe you passed through is here
super().__init__()
# Frame for TreeView
frame1 = tk.LabelFrame(self, text="Excel Data")
frame1.pack(fill="both", expand="true")
frame1.pack_propagate(0)
# the size of the window.
self.geometry("500x500")
self.resizable(0, 0)
self.title("Your Excel File in Tkinter") # the window title
# This creates your Treeview widget.
tv1 = ttk.Treeview(frame1)
tv1.place(relheight=1, relwidth=1) # set the height and width of the widget to 100% of its container (frame1).
treescrolly = tk.Scrollbar(frame1, orient="vertical", command=tv1.yview) # command means update the yaxis view of the widget
treescrollx = tk.Scrollbar(frame1, orient="horizontal", command=tv1.xview) # command means update the xaxis view of the widget
tv1.configure(xscrollcommand=treescrollx.set, yscrollcommand=treescrolly.set) # assign the scrollbars to the Treeview Widget
treescrollx.pack(side="bottom", fill="x") # make the scrollbar fill the x axis of the Treeview widget
treescrolly.pack(side="right", fill="y") # make the scrollbar fill the y axis of the Treeview widget
# this loads the dataframe into the treeview widget
tv1["column"] = list(your_dataframe.columns)
tv1["show"] = "headings"
for column in tv1["columns"]:
tv1.heading(column, text=column) # let the column heading = column name
df_rows = your_dataframe.to_numpy().tolist() # turns the dataframe into a list of lists
for row in df_rows:
tv1.insert("", "end", values=row) # inserts each list into the treeview. For parameters see https://docs.python.org/3/library/tkinter.ttk.html#tkinter.ttk.Treeview.insert
root.mainloop()
我通过结合来自 Whosebug 的一些答案自己弄明白了。 基本上通过一个函数而不是我的“普通”框架创建一个新的 window，然后使用 command=DisplayDF 创建一个按钮。这确保它只看到最新的数据帧，并且在调用时不会因为在导入之前被引用而产生错误。
def DisplayDF():
global df
# create child window
win = Toplevel()
# display message
message = "Current output (Max 100 rows)"
tk.Label(win, text=message).pack()
text = tk.Text(win)
text.insert(tk.END, str(df.head(100)))
text.pack()