对函数中的变量使用 Tkinter 用户输入答案

【问题标题】：Using Tkinter user input for variables in functions对函数中的变量使用 Tkinter 用户输入
【发布时间】：2015-12-14 14:11:26
【问题描述】：

我有两个功能。我的第一个函数创建了一个 GUI，用户可以在其中输入 8 个不同物种的最小值和最大值。我的第二个函数尝试使用这些最小值和最大值在它们各自的最小值和最大值的边界内创建 1000 种混合物的模拟，同时遵守许多不同的约束。但是，当我运行模拟时，我没有得到任何值。我只得到带有物种标题的 CSV 文件。我也没有得到任何有价值的错误。我的代码在下面，我不知道如何完成这项工作。任何帮助将非常感激。

import Tkinter 
import pandas as pd
import numpy as np

class simulation_tk(Tkinter.Tk):
    def __init__(self,parent):
        Tkinter.Tk.__init__(self,parent)
        self.parent = parent
        self.initialize()
        self.grid()

    def initialize(self):

        self.c2_low =Tkinter.StringVar()
        self.c3_low =Tkinter.StringVar()
        self.ic4_low =Tkinter.StringVar()
        self.nc4_low =Tkinter.StringVar()
        self.ic5_low =Tkinter.StringVar()
        self.nc5_low =Tkinter.StringVar()
        self.neoc5_low =Tkinter.StringVar()
        self.n2_low = Tkinter.StringVar()

        self.c2_high =Tkinter.StringVar()
        self.c3_high =Tkinter.StringVar()
        self.ic4_high =Tkinter.StringVar()
        self.nc4_high =Tkinter.StringVar()
        self.ic5_high =Tkinter.StringVar()
        self.nc5_high =Tkinter.StringVar()
        self.neoc5_high=Tkinter.StringVar()
        self.n2_high = Tkinter.StringVar()

        self.entry = Tkinter.Entry(self, textvariable = self.c2_low).grid(column=0,row=1,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.c2_high).grid(column=0,row=2,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.c3_low).grid(column=0,row=3,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.c3_high).grid(column=0,row=4,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic4_low).grid(column=1,row=1,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic4_high).grid(column=1,row=2,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc4_low).grid(column=1,row=3,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc4_high).grid(column=1,row=4,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic5_low).grid(column=0,row=5,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic5_high).grid(column=0,row=6,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc5_low).grid(column=0,row=7,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc5_high).grid(column=0,row=8,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.neoc5_low).grid(column=1,row=5,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.neoc5_high).grid(column=1,row=6,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.n2_low).grid(column=1,row=7,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.n2_high).grid(column=1,row=8,sticky='EW')

        self.resizable(False,False)


        button = Tkinter.Button(self,text=u"simulate", command =self.simulation)
        button.grid(column=3,row=9)

    def simulation(self):


        sample_runs =10000     # Sample Population needs to be higher than exporting population
        export_runs = 1000    # How many samples we actually take

        c2_low = self.c2_low.get()
        c2_high = self.c2_high.get()
        c3_low = self.c3_low.get()
        c3_high = self.c3_high.get()
        ic4_low = self.ic4_low.get()
        ic4_high =self.ic4_high.get()
        nc4_low =self.nc4_low.get()
        nc4_high = self.nc4_high.get()  
        ic5_low = self.ic5_low.get()
        ic5_high = self.ic5_high.get()
        nc5_low = self.nc5_low.get()
        nc5_high = self.nc5_high.get()
        neoc5_low = self.neoc5_low.get()
        neoc5_high = self.neoc5_high.get()
        n2_low = self.n2_low.get()
        n2_high = self.n2_high.get()

        c2 = np.random.uniform(c2_low,c2_high,sample_runs)
        c3 = np.random.uniform(c3_low,c3_high, sample_runs)
        ic4 = np.random.uniform(ic4_low,ic4_high,sample_runs)
        nc4 = np.random.uniform(nc4_low,nc4_high,sample_runs)
        ic5 = np.random.uniform(ic5_low,ic5_high,sample_runs)
        nc5 = np.random.uniform(nc5_low,nc5_high,sample_runs)
        neoc5 = np.random.uniform(neoc5_low ,neoc5_high,sample_runs)
        n2 = np.random.uniform(n2_low, n2_high,sample_runs)

        # SETS CONSTRAINTS BASED ON RANGES

        masked = np.where((c3>=c3_low) & (c3<=c3_high) & (c2>=c2_low) & (c2<= c2_high) & (ic4>=ic4_low) & 
        (ic4<= ic4_high) & (nc4>= nc4_low) & (nc4<= nc4_high) & (ic5>= ic5_low) & (ic5<= ic5_high)& (nc5>= nc5_low)& 
        (nc5<= nc5_high)& (neoc5>= neoc5_low)& (neoc5<=neoc5_high) & (n2>=n2_low) & (n2<= n2_high))

        # MASKED CREATES AN INDEX (Where constraints are held) FOR LOOKING THROUGH DATA

        c2 = c2[masked][:export_runs]
        c3 = c3[masked][:export_runs]
        ic4 = ic4[masked][:export_runs]
        nc4 = nc4[masked][:export_runs]
        ic5 = ic5[masked][:export_runs]
        nc5 = nc5[masked][:export_runs]
        neoc5 = neoc5[masked][:export_runs]
        n2 = n2[masked][:export_runs]

        # DETERMINES CONC FROM METHANE BY BALANCE

        c1 = 100-c2-c3-nc4-ic4-nc5-ic5-neoc5-n2

        #CREATES A SERIES FOR EACH COMPONENET AND ADDS COLUMNS TO A FINAL DATAFRAME 

        c1_ser = pd.Series(c1)
        c2_ser = pd.Series(c2)
        c3_ser = pd.Series(c3)
        ic4_ser = pd.Series(ic4)
        nc4_ser = pd.Series(nc4)
        ic5_ser = pd.Series(ic5)
        nc5_ser = pd.Series(nc5)
        neoc5_ser = pd.Series(neoc5)
        n2_ser = pd.Series(n2)


        #EXPORTS DATAFRAME TO .CSV FILE NAMED LNG_DATA

        df = pd.DataFrame([c1_ser, c2_ser, c3_ser, ic4_ser, nc4_ser, ic5_ser, nc5_ser, neoc5_ser, n2_ser]).T
        df.columns = ['C1','C2','C3','nC4','iC4','nC5','iC5','neoC5','N2']
        df.to_csv(path to directory you want the saved file)

if __name__ == "__main__":
    app = simulation_tk(None)
    app.title('Simulation')
    app.mainloop()

编辑：

原模拟函数代码如下：

import numpy as np
import pandas as pd
import time

def LNG_SIMULATION(no_of_simulations):

    t0 = time.time()

    # SET COMPOSITION RANGES HERE:

    c2_low    =0;    c2_high   =14 
    c3_low    =0;    c3_high   =4
    nc4_low   =0;   nc4_high   =1.5
    ic4_low   =0;   ic4_high   =1.2
    nc5_low   =0;   nc5_high   =0.1
    ic5_low   =0;   ic5_high   =0.1
    neoc5_low =0; neoc5_high   =0.01
    n2_low    =0;    n2_high   =1.5

    # PRODUCES A RANDOM UNIFORM DISTRIBUTION BETWEEN LOW AND HIGH * runs

    sample_runs =10000     # Sample Population needs to be higher than exporting population
    export_runs = no_of_simulations    # How many samples we actually take

    c2 = np.random.uniform(c2_low,c2_high,sample_runs)
    c3 = np.random.uniform(c3_low,c3_high, sample_runs)
    ic4 = np.random.uniform(ic4_low,ic4_high,sample_runs)
    nc4 = np.random.uniform(nc4_low,nc4_high,sample_runs)
    ic5 = np.random.uniform(ic5_low,ic5_high,sample_runs)
    nc5 = np.random.uniform(nc5_low,nc5_high,sample_runs)
    neoc5 = np.random.uniform(neoc5_low,neoc5_high,sample_runs)
    n2 = np.random.uniform(n2_low, n2_high,sample_runs)

    # SETS CONSTRAINTS BASED ON RANGES

    masked = np.where((c3>=0) & (c3<=4) & (c2>=0) & (c2<=14) & (ic4>=0) & 
    (ic4<=1.5) & (nc4>=0) & (nc4<=1.2) & (ic5>=0) & (ic5<=0.1)& (nc5>=0)& 
    (nc5<=0.1)& (neoc5>=0)& (neoc5<=0.01) & (n2>=0) & (n2<=1.5))

    # MASKED CREATES AN INDEX (Where constraints are held) FOR LOOKING THROUGH DATA

    c2 = c2[masked][:export_runs]
    c3 = c3[masked][:export_runs]
    ic4 = ic4[masked][:export_runs]
    nc4 = nc4[masked][:export_runs]
    ic5 = ic5[masked][:export_runs]
    nc5 = nc5[masked][:export_runs]
    neoc5 = neoc5[masked][:export_runs]
    n2 = n2[masked][:export_runs]

    # DETERMINES CONC FROM METHANE BY BALANCE

    c1 = 100-c2-c3-nc4-ic4-nc5-ic5-neoc5-n2

    #CREATES A SERIES FOR EACH COMPONENET AND ADDS COLUMNS TO A FINAL DATAFRAME 

    c1_ser = pd.Series(c1)
    c2_ser = pd.Series(c2)
    c3_ser = pd.Series(c3)
    ic4_ser = pd.Series(ic4)
    nc4_ser = pd.Series(nc4)
    ic5_ser = pd.Series(ic5)
    nc5_ser = pd.Series(nc5)
    neoc5_ser = pd.Series(neoc5)
    n2_ser = pd.Series(n2)

    print np.min(c1); print np.max(c1) # Check for methane range

    #EXPORTS DATAFRAME TO .CSV FILE NAMED LNG_DATA

    df = pd.DataFrame([c1_ser, c2_ser, c3_ser, ic4_ser, nc4_ser, ic5_ser, nc5_ser, neoc5_ser, n2_ser]).T
    df.columns = ['C1','C2','C3','nC4','iC4','nC5','iC5','neoC5','N2']
    df.to_csv(filepath)

    t1 = time.time()
    tfinal = t1-t0, 'seconds'
    print tfinal


LNG_SIMULATION(1000)

这会以 csv 文件的形式提供以下输出：

每行加起来为 100，因此 c1 = 100-（所有其他分量的总和）

C1  C2  C3  nC4 iC4 nC5 iC5 neoC5   N2
0   82.85372539 12.99851014 2.642744858 0.129878248 0.800397967 0.002835756 0.01996335  0.00665644  0.545287856
1   97.53896049 1.246468861 0.00840227  0.616819596 0.340552181 0.093463733 0.0415282   0.002044789 0.11175988
2   96.06680372 1.005440722 0.427965685 0.944281965 0.354424967 0.029694142 0.046906668 0.001961002 1.122521133
3   92.152083   4.558717345 1.850648013 0.060053009 0.802721707 0.055533032 0.013490485 0.008897805 0.497855601
4   81.68486996 13.21690811 2.478113198 0.825638261 0.963227282 0.02162254  0.03812538  0.006329348 0.765165918
5   86.4237313  9.387647074 2.729233511 0.562534986 0.786110737 0.050537327 0.026122606 0.000290321 0.033792141
6   95.11319788 2.403944121 0.467770537 0.229967177 0.220494035 0.073742963 0.007893607 0.007473005 1.475516673
7   92.501114   2.677293658 2.742409857 0.608661787 0.237898432 0.073326044 0.030292277 0.002908029 1.126095919
8   89.83876672 5.850123215 2.598266005 0.060712896 0.29401403  0.037017143 0.048577495 0.001888549 1.270633946
9   84.14677099 13.9234657  0.214404288 0.535574576 0.677735065 0.061556983 0.015255684 0.006789481 0.418447232
10  94.73390493 2.302821233 1.478361587 0.500991046 0.022823156 0.030764131 0.024351373 0.009064709 0.896917832

1000 行。

最终编辑：

        self.entry = Tkinter.Entry(self, textvariable = self.c2_low).grid(column=0,row=1,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.c2_high).grid(column=1,row=1,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.c3_low).grid(column=0,row=2,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.c3_high).grid(column=1,row=2,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic4_low).grid(column=0,row=3,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic4_high).grid(column=1,row=3,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc4_low).grid(column=0,row=4,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc4_high).grid(column=1,row=4,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic5_low).grid(column=0,row=5,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic5_high).grid(column=1,row=5,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc5_low).grid(column=0,row=6,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc5_high).grid(column=1,row=6,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.neoc5_low).grid(column=0,row=7,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.neoc5_high).grid(column=1,row=7,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.n2_low).grid(column=0,row=8,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.n2_high).grid(column=1,row=8,sticky='EW')

【问题讨论】：

注意：我知道模拟是有效的，因为我已经在类函数之外单独运行它。
“我没有价值”的字面意思是什么？试图获取值的代码是否失败？它是否成功但值为None？它是否成功但值为空字符串？您是否可以从 GUI 获取值，但您的计算没有返回任何结果？
好的，我的意思是 CSV 文件中的单元格是空的。标题存在，但单元格为空。我知道模拟工作“没有 GUI 用户输入”，所以问题必须在获取用户输入数据和尝试在模拟中使用之间的某个地方。
你做了什么调试？第一步应该始终是验证您的假设。打印出从 GUI 获得的值，然后再将它们输入函数。它们是您期望的值吗？它们是正确的数据类型吗？

标签： python pandas tkinter tkinter-canvas

【解决方案1】：

问题在于，在您的 np.where 调用中，您的比较是在字符串值（即 c2_low、c2_high 等中的值）和 numpy 数组之间执行的。这种比较是行不通的。您需要将这些字符串转换为浮点数，如下所示：

c2_low = float(self.c2_low.get())

我还要注意，我认为您不需要致电 np.where。您所做的只是确保c2、c3 等的值都在指定范围内。默认情况下应该是这样；当您调用np.random.uniform 时，这些数组就是这样设置的。因此，您应该能够完全取消您的 masked 变量。如果我对您的代码进行这些更改，我将得到以下结果：

import Tkinter as Tkinter
import pandas as pd
import numpy as np

class simulation_tk(Tkinter.Tk):
    def __init__(self,parent):
        Tkinter.Tk.__init__(self,parent)
        self.parent = parent
        self.initialize()
        self.grid()

    def initialize(self):

        self.c2_low =Tkinter.StringVar()
        self.c3_low =Tkinter.StringVar()
        self.ic4_low =Tkinter.StringVar()
        self.nc4_low =Tkinter.StringVar()
        self.ic5_low =Tkinter.StringVar()
        self.nc5_low =Tkinter.StringVar()
        self.neoc5_low =Tkinter.StringVar()
        self.n2_low = Tkinter.StringVar()

        self.c2_high =Tkinter.StringVar()
        self.c3_high =Tkinter.StringVar()
        self.ic4_high =Tkinter.StringVar()
        self.nc4_high =Tkinter.StringVar()
        self.ic5_high =Tkinter.StringVar()
        self.nc5_high =Tkinter.StringVar()
        self.neoc5_high=Tkinter.StringVar()
        self.n2_high = Tkinter.StringVar()

        self.entry = Tkinter.Entry(self, textvariable = self.c2_low).grid(column=0,row=1,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.c2_high).grid(column=0,row=2,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.c3_low).grid(column=0,row=3,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.c3_high).grid(column=0,row=4,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic4_low).grid(column=1,row=1,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic4_high).grid(column=1,row=2,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc4_low).grid(column=1,row=3,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc4_high).grid(column=1,row=4,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic5_low).grid(column=0,row=5,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.ic5_high).grid(column=0,row=6,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc5_low).grid(column=0,row=7,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.nc5_high).grid(column=0,row=8,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.neoc5_low).grid(column=1,row=5,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.neoc5_high).grid(column=1,row=6,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.n2_low).grid(column=1,row=7,sticky='EW')
        self.entry = Tkinter.Entry(self, textvariable = self.n2_high).grid(column=1,row=8,sticky='EW')

        self.resizable(False,False)


        button = Tkinter.Button(self,text=u"simulate", command =self.simulation)
        button.grid(column=3,row=9)

    def simulation(self):


        sample_runs =10000     # Sample Population needs to be higher than exporting population
        export_runs = 1000    # How many samples we actually take

        c2_low = float(self.c2_low.get())
        c2_high = float(self.c2_high.get())
        c3_low = float(self.c3_low.get())
        c3_high = float(self.c3_high.get())
        ic4_low = float(self.ic4_low.get())
        ic4_high = float(self.ic4_high.get())
        nc4_low = float(self.nc4_low.get())
        nc4_high = float(self.nc4_high.get())
        ic5_low = float(self.ic5_low.get())
        ic5_high = float(self.ic5_high.get())
        nc5_low = float(self.nc5_low.get())
        nc5_high = float(self.nc5_high.get())
        neoc5_low = float(self.neoc5_low.get())
        neoc5_high = float(self.neoc5_high.get())
        n2_low = float(self.n2_low.get())
        n2_high = float(self.n2_high.get())

        c2 = np.random.uniform(c2_low,c2_high,sample_runs)
        c3 = np.random.uniform(c3_low,c3_high, sample_runs)
        ic4 = np.random.uniform(ic4_low,ic4_high,sample_runs)
        nc4 = np.random.uniform(nc4_low,nc4_high,sample_runs)
        ic5 = np.random.uniform(ic5_low,ic5_high,sample_runs)
        nc5 = np.random.uniform(nc5_low,nc5_high,sample_runs)
        neoc5 = np.random.uniform(neoc5_low ,neoc5_high,sample_runs)
        n2 = np.random.uniform(n2_low, n2_high,sample_runs)

        # SETS CONSTRAINTS BASED ON RANGES

        # masked = np.where((c3>=c3_low) & (c3<=c3_high) & (c2>=c2_low) & (c2<= c2_high) & (ic4>=ic4_low) &
        #                   (ic4<= ic4_high) & (nc4>= nc4_low) & (nc4<= nc4_high) & (ic5>= ic5_low) & (ic5<= ic5_high)& (nc5>= nc5_low)&
        #                   (nc5<= nc5_high)& (neoc5>= neoc5_low)& (neoc5<=neoc5_high) & (n2>=n2_low) & (n2<= n2_high))

        # MASKED CREATES AN INDEX (Where constraints are held) FOR LOOKING THROUGH DATA

        c2 = c2[:export_runs]
        c3 = c3[:export_runs]
        ic4 = ic4[:export_runs]
        nc4 = nc4[:export_runs]
        ic5 = ic5[:export_runs]
        nc5 = nc5[:export_runs]
        neoc5 = neoc5[:export_runs]
        n2 = n2[:export_runs]

        # DETERMINES CONC FROM METHANE BY BALANCE

        c1 = 100-c2-c3-nc4-ic4-nc5-ic5-neoc5-n2

        #CREATES A SERIES FOR EACH COMPONENET AND ADDS COLUMNS TO A FINAL DATAFRAME

        c1_ser = pd.Series(c1)
        c2_ser = pd.Series(c2)
        c3_ser = pd.Series(c3)
        ic4_ser = pd.Series(ic4)
        nc4_ser = pd.Series(nc4)
        ic5_ser = pd.Series(ic5)
        nc5_ser = pd.Series(nc5)
        neoc5_ser = pd.Series(neoc5)
        n2_ser = pd.Series(n2)


        #EXPORTS DATAFRAME TO .CSV FILE NAMED LNG_DATA

        df = pd.DataFrame([c1_ser, c2_ser, c3_ser, ic4_ser, nc4_ser, ic5_ser, nc5_ser, neoc5_ser, n2_ser]).T
        df.columns = ['C1','C2','C3','nC4','iC4','nC5','iC5','neoC5','N2']
        df.to_csv('output.csv')

if __name__ == "__main__":
    app = simulation_tk(None)
    app.title('Simulation')
    app.mainloop()

我已经使用 Python 2.7 和 numpy 1.7.1 以及带有 numpy 1.9.2 的 Python 3.4（对 tkinter 导入语句进行了适当更改）对此进行了测试，在这两种情况下，我都得到了一个完全填充的 CSV 文件，其中每个行总和为 100。

【讨论】：

不幸的是，这不起作用。我的 CSV 文件中仍然有空单元格。
您能否编辑您的原始帖子以描述您用作输入的值？在为第 61-76 行添加强制转换并在 Python 2.7 中运行它后，我得到一个填充的 CSV 文件，其中包含 1001 行和 10 列（我不能说它们是否正确）。我使用 10 和 20 的交替值作为输入（即c2_low = 10、c2_high = 20、c3_low = 10...）
我在我的答案中添加了额外的细节，以及我正在运行的源代码的副本。你运行的是什么版本的 Python/numpy？如果您在使用c2_low、c2_high 创建数组之前检查它们的值，这些值是您所期望的吗？数组c1、c2...呢？这些是否包含您所期望的？我要回应上面的@BryanOakley：花一些时间在调试器（甚至打印语句）上会得到回报。
我已经运行了你提出的代码。但是，输出不是我想要的。 'masked' 部分确保满足约束条件。这很难解释。但是，如果你看到我的编辑帖子，你可以看到所有数字都是非零和随机的，但所有数字都加到 100。
按照@BryanOakley 的建议，通过一些调试解决了问题。问题是条目被分配到错误的标签。我已经编辑了我的帖子以获得正确的代码。谢谢。