StackGP.printGPModel

StackGP.printGPModel(model,inputData=symbols([“x”+str(i) for i in range(100)]))

printGPModel is a StackGP function that evaluates a GP model with the supplied data.

The function expects 1 arguments: model

There is 1 optional argument: inputData

The arguments are described below:

• model: A StackGP model.

• inputData: a list of SymPy symbols to use as the variable names.

---

First we need to load in the necessary packages

import StackGP as sgp
import numpy as np

Overview

Display a model

Here we generate a random model with up to 4 variables, the default operator set, the default constant set, and a maxSize of 10.

randomModel=sgp.generateRandomModel(4, sgp.defaultOps(), sgp.defaultConst(), 10)

We can display the random model below

sgp.printGPModel(randomModel)

\[\displaystyle e^{2 \sqrt{x_{3}}}\]

---

---

Options

This section showcases how each of the different option settings can be used with the printGPModel function.

---

inputData

inputData can be used to change the symbols used when displaying the model. The default is to use x1, x2, …, x100

First lets generate some random data to fit a model to.

#Define a challenging function to generate data
def demoFunc(x,y):
    return np.sin(x) + y
#Generate data
inputData=np.array([np.random.randint(1,10,10),np.random.randint(1,10,10)])
response=demoFunc(inputData[0],inputData[1])

Now lets try fitting with the default settings. We will turn on tracking so we can see the progress.

#Generate models
models=sgp.evolve(inputData,response,liveTracking=True, generations=100, popSize=100, ops=sgp.allOps())

Now we can display the best model evolved.

#View best model
sgp.printGPModel(models[0])

\[\displaystyle 0.999999999999999 x_{1} + 0.999999999999999 \sin{\left(x_{0} \right)} + 5.61733354972272 \cdot 10^{-16}\]

Now maybe we don’t want x and y as the variable names and instead want something like var1 and var2. We can do this using the inputData argument like below.

from sympy import symbols
sgp.printGPModel(models[0], inputData=symbols("var1,var2"))

\[\displaystyle 0.999999999999999 var_{2} + 0.999999999999999 \sin{\left(var_{1} \right)} + 5.61733354972272 \cdot 10^{-16}\]

---

---

Examples

This section showcases how each of the different arguments can be used with the printGPModel function.

---

Checking the form of a random model using specific variable names

randomModel = sgp.generateRandomModel(4, sgp.defaultOps(), sgp.defaultConst(), 10)
sgp.printGPModel(randomModel, inputData=symbols("temperature , pressure, humidity, windSpeed"))

\[\displaystyle humidity + \frac{e^{2}}{pressure}\]

Checking the forms of models from an evolved Pareto front

First we generate some random data and evolve a model population.

inputData, responseData, model = sgp.generateRandomBenchmark()
models = sgp.evolve(inputData,responseData,liveTracking=True, generations=100, popSize=100, ops=sgp.allOps())

Now we can grab the Pareto front of the models

pFront = sgp.selectModels(models, selectionSize=2)

for i, model in zip(range(len(pFront)), pFront):
    print(i, sgp.printGPModel(model, inputData=symbols("position, velocity, acceleration, time, mass")))

0 -3.5527136788005e-16 + 1.0/velocity
1 7.35939390871629 - 8.29633610621588*velocity