Question

我有一个pyomo模型，其可变参数为“ model.Prices”。使用for循环，该模型能够迭代并找到最佳解决方案。但是我不知道如何将所有迭代的解决方案（包括参数，目标，变量的值）写入excel / csv文件。

这是for循环

for RapeseedPrice in range(300, 600, 100):
    # change the value of parameter model.Prices
    model.Prices["rapeseed"] = RapeseedPrice
    print("When price of rapeseed = {}".format(RapeseedPrice))
    opt = SolverFactory('ipopt')
    opt.solve(model, tee = True) 
    results = opt.solve
    model.display()

如果我包括以下三个for循环，则可以访问所需的所有值：

for RapeseedPrice in range(300, 600, 100):
    # change the value of parameter model.Prices
    model.Prices["rapeseed"] = RapeseedPrice
    print("When price of rapeseed = {}".format(RapeseedPrice))
    opt = SolverFactory('ipopt')
    opt.solve(model, tee = True) 
    results = opt.solve
    model.display()
    for parmobject in model.component_objects(Param, active=True):
        nametoprint = str(str(parmobject.name))
        print ("Parameter ", nametoprint) # doctest: +SKIP
        for index in parmobject:
            vtoprint = value(parmobject[index])
            print (" ",index, vtoprint) # doctest: +SKIP
    for o in model.component_data_objects(Objective, active=True):
        print(o, value(o)) 
    for v in model.component_data_objects(Var, active=True):
        print(v, value(v))

我希望一个excel文件具有不同的交互作用（作为列），所有参数，目标，变量作为行。

            it1        it2       it3
param 1
param 2
param 3
...
objective
var 1
var 2
var 3
...

Answer 1

这样的事情应该可以将您的值放入Pandas DataFrame中，并从那里很容易输出到csv文件：

from pyomo.environ import *
import pandas as pd

m = ConcreteModel()

m.s = Set(initialize=[1,2,3])
m.p = Param(initialize=1, mutable=True)
m.x = Var(m.s, bounds=(1,3))
m.obj = Objective(expr=m.p*sum(m.x[k]**2 for k in m.s))

solver = SolverFactory('ipopt')

all_data={}
for j in range(1,4):

    m.p = j
    solver.solve(m)

    data = {}

    for i in m.component_data_objects(Param):
        data[i.name] = value(i)
    for i in m.component_data_objects(Var):
        data[i.name] = value(i)
    for i in m.component_data_objects(Objective):
        data[i.name] = value(i)

    all_data['Solve '+str(j)] = pd.Series(data)

df = pd.DataFrame(all_data)

Answer 2

让我回答我自己的问题。

为此，我们需要导入：

import pandas as pd
import numpy as np
from copy import deepcopy

首先，您需要获取标题：

header0 = []
header1 = []

for parmobject in model.component_objects(Param, active=True):
        nametoprint = str(str(parmobject.name))        
        for index in parmobject:
                header0.append(nametoprint)
                header1.append(index)
for o in model.component_data_objects(Objective, active=True):
        header0.append("Objective")
        header1.append(str(o.name))
for v in model.component_data_objects(Var, active=True):
        header0.append("Variable")
        header1.append(str(v.name))

MultiHeaders = [header0, header1]

第二，您必须获取值：

AllData = []
pov_data = []

for RapeseedPrice in range(100, 500, 100): 

    model.Prices['rapeseed'] = RapeseedPrice

    # solve the model
    opt.solve(model) 

    # access pov_dat  
    for parmobject in model.component_objects(Param, active=True):
            for index in parmobject:
                    vtoprint = value(parmobject[index])
                    pov_data.append(vtoprint)
    for o in model.component_data_objects(Objective, active=True):
            pov_data.append(value(o))

    for v in model.component_data_objects(Var, active=True):
            pov_data.append(value(v)) 

    AllData.append(deepcopy(pov_data))
    pov_data.clear()

第三，您将数据和标头合并到一个数据框中：

CDFarm_results = pd.DataFrame(data = np.array(AllData), columns = 
     MultiHeaders)

最后，您将数据框保存到excel文件中。

writer = pd.ExcelWriter('CDFarm_results_pandas.xlsx', engine='xlsxwriter')
CDFarm_results.to_excel(writer, 'Sheet1')
writer.save()

请在这里查看要点：该脚本能够将所有参数，目标和变量的名称和值保存到excel文件my-gist-link

中

最终的excel看起来像这样：

如何将多次迭代的pyomo解决方案（仅包括参数，目标和变量）写入excel / csv文件？

2 个答案: