Exercise 1

In this exercise, you will try to use FMT to read a scenario, solve it, read one of its outputs, and export the solution.

In order to do the exercise, you will need to have FMT installed on your computer, and to have downloaded the example files from the GitHub repository of FMT.

💾 Downloading and installing FMT

💾 Downloading the example files of FMT

Goals of the exercise

Use FMT to :

Read the Woodstock model TWD_land in the example files by selecting its scenario named LP.
- Remember to select the CLP solver while parsing the model.
Solve the model by
- Constructing its full graph for 10 periods
- Setting the objective and constraints of the model
- Using the FMTmodel.initialsolve() function to solve it.
Read the values of the output OSUPREC of the model by
- Getting a list of the outputs of the model with the function FMTmodel.getouputs()
- Doing a for loop to select the right output with the FMToutput.getname() function
- Printing the output values using the function FMTmodel.getvariabilities(FMToutput)
Write the optimized schedule of the model in a .txt file by
- Creating a new FMTscheduleparser object
- Getting the solution of the model at each period with a list using LPModel.getsolution(period)
- Transforming the list into a vector
- Writing the vector into a .txt file with FMTscheduleparser.write()

You can use the information from the previous pages of documentation to help you.

When you are done, or if you are stuck, you can display the correction of this exercise for Python by clicking on the “Reveal correction” below.

Correction

Reveal correction

#%% LOADING FMT

# Here, we load FMT directly, as if it were installed
# via pip.

from FMT import Models
from FMT import Parser
from FMT import Version

#%% LOADING THE WOODSTOCK MODEL

# Creation of a parser to read the model
modelParser = Parser.FMTmodelparser()

# Defining the path leading to the Woodstock model .pri file.
# WARNING : try to have a path without spaces inside it.
pathToWoodstockModel = "<Path to example folder>/Examples/Models/TWD_land/TWD_land.pri"

# Defining the scenario to be read, which contains a set of constraints and objectives.
scenariosToSelect = ["LP"]

# Reading the model using the parser and the scenario name. We can read multiple at once,
# but here, we'll only use one.
listOfModelsParsed = modelParser.readproject(pathToWoodstockModel, scenariosToSelect)

# Loading the model from the list of models given by the parser. We only take the first
# one, which is the only one that we want. We also indicate the solver interface that we
# will use to solve the linear programming model; here, we will use CLP, the free and
# open-source solver.
LPModel = Models.FMTlpmodel(listOfModelsParsed[0], Models.FMTsolverinterface.CLP)

#%% SOLVING THE MODEL

# We build the full graph of the model for 10 periods.
for period in range(1, 11):
    print(LPModel.buildperiod())

# The objective is simply the first item of the
# constraints list, that we take away from the list with pop().
constraintsOfModel = LPModel.getconstraints()
objectiveOfModel = constraintsOfModel.pop(0)

# We set the contraints and objectives of the model.
for constraint in constraintsOfModel:
    print(LPModel.setconstraint(constraint))
print(LPModel.setobjective(objectiveOfModel))

# We solve the model.
LPModel.initialsolve()

#%% READING THE OSUPREC OUTPUT

# We can read OSUPREC before the model is solved, but the values won't be interesting
# to us.
outputWanted = "OSUPREC"

# We get the outputs of the model
outputsOfModel = LPModel.getoutputs()
# We loop around the list of output to select the one with the
# right name
for output in outputsOfModel:
    print(output.getname())
    if output.getname() == outputWanted:
    	# We print the measures of "OSUPREC" in the python prompt
        print(LPModel.getvariabilities([output]))
        # We save the values of OSUPREC to format them better
        osuprecValues = LPModel.getvariabilities([output])
       
# Here is a small code to display the values better,
# adapting to the fact that the getvariabilities() function
# returns the minimum, mean and maximum value of the output.
print("\n\n ###############################")
print("Values of " + outputWanted + " :\n")
print("PERIOD\t\tMINIMUM\t\tMEAN\t\tMAXIMUM")
for period in range(1, 11):
    print(str(period) + "\t\t\t" + str(round(osuprecValues["L"+outputWanted][period-1], 2))
          + "\t\t" + str(round(osuprecValues["M"+outputWanted][period-1], 2))
          + "\t\t" + str(round(osuprecValues["U"+outputWanted][period-1], 2)))

#%% WRITING THE SCHEDULE

# We create a new scheduleparser object
scheduleParser = Parser.FMTscheduleparser()

# We get the solution of the model
modelSolution = list()
for period in range(1, 11):
    modelSolution.append(LPModel.getsolution(period))
    
# We write everything into a .txt file after putting everything as a vector.
modelSolutionAsVector = [[i] for i in modelSolution]
scheduleParser.write(modelSolution, r"D:/TempCode/FMT/FMT/Examples/programme.txt")

Last updated on May 5, 2019