This folder contains the STOCK PRICE SIMULATOR DEVS model implemented in Cadmium and Keras/TensorFlow
/**************************/
/****FILES ORGANIZATION****/
/**************************/
README.txt
SPS.docx
makefile
machine_learning [this folder contains the files to train a deep learning model]
GOOG.svc [last 5 years data from google stock price]
train.py []
scaler.scale [scale file]
trained_model.h5 [exported trained model]
atomics [This folder contains atomic models implemented in Cadmium]
prediction_generator.hpp
accuracy_evaluator.hpp
bin [This folder will be created automatically the first time you compile the project.
It will contain all the executables]
build [This folder will be created automatically the first time you compile the project.
It will contain all the build files (.o) generated during compilation]
data_structures [This folder contains message data structure used in the model]
time_series.hpp
time_series.cpp
input_data [This folder contains all the input data to run the model and the tests]
prediction_generator_input_test1.txt
prediction_generator_input_test2.txt
prediction_generator_input_test2.txt
accuracy_evaluator_input_test1.txt
accuracy_evaluator_input_test2.txt
accuracy_evaluator_input_test2.txt
stock_prediction_simulation_input_test.txt
simulation_results [This folder will be created automatically the first time you compile the project.
It will store the outputs from your simulations and tests]
test [This folder the unit test of the atomic models]
main_prediction_generator_test.cpp
main_accuracy_evaluator_test.cpp
top_model [This folder contains the Asset Price Simulator top model]
main_.cpp
scripts
[This folder contains the scripts to execute the bin files with the different input_data files]
stock_prediction_simulation.sh [executes STOCK_PREDICTION_SIMULATION file (compiled in the bin folder) with the stock_prediction_simulation_input_test.txt input data file]
prediction_generator_test1.sh [executes PREDICTION_GENERATOR_TEST file (compiled in the bin folder) with the prediction_generator_input_test1.txt input data file]
prediction_generator_test2.sh [executes PREDICTION_GENERATOR_TEST file (compiled in the bin folder) with the prediction_generator_input_test2.txt input data file]
prediction_generator_test3.sh [executes PREDICTION_GENERATOR_TEST file (compiled in the bin folder) with the prediction_generator_input_test3.txt input data file]
accuracy_evaluator_test1.sh [executes ACCURACY_EVALUATOR_TEST file (compiled in the bin folder) with the accuracy_evaluator_input_test1.txt input data file]
accuracy_evaluator_test2.sh [executes ACCURACY_EVALUATOR_TEST file (compiled in the bin folder) with the accuracy_evaluator_input_test2.txt input data file]
accuracy_evaluator_test3.sh [executes ACCURACY_EVALUATOR_TEST file (compiled in the bin folder) with the accuracy_evaluator_input_test3.txt input data file]
scaler.scale [scale file required by tensorFlow]
predictor.py [contains the function that calls the trained model]
trained_model.h5 [deep learning trained model]
/*************/
/****STEPS****/
/*************/
0 - SPS.docx contains the explanation of this model
1 - Install Cadmium and dependencies.
2. Update include path in the makefile in this folder and subfolders. You need to update the following lines:
INCLUDECADMIUM=-I ../../cadmium/include
INCLUDEDESTIMES=-I ../../DESTimes/include
Update the relative path to cadmium/include from the folder where the makefile is. You need to take into account where you copied the folder during the installation process
Example: INCLUDECADMIUM=-I ../../cadmium/include
Do the same for the DESTimes library
NOTE: if you follow the step by step installation guide you will not need to update these paths.
3. Install TensorFlow and Keras
1. open a terminal in the main folder.
2. Install the machine learning dependencies, type in the terminal "sudo ./install_dependencies.sh"
2 - Compile the project and the tests
1 - Open the terminal (Ubuntu terminal for Linux and Cygwin for Windows) in the APS folder
2 - To compile the project and the tests, type in the terminal:
make clean; make all
3 - Run tests
1 - Open the terminal in the scripts folder.
2 - To run the test, type in the terminal "./NAME_OF_THE_SCRIPT.sh" (For windows, inside the script the executable should be changed to the extension .exe).
3 - To check the output of the test, go to the folder simulation_results.
4 - Run the top model
1 - Open the terminal (Ubuntu terminal for Linux and Cygwin for Windows) in the scripts folder.
2 - To run the model, type in the terminal "./stock_prediction_simulator.sh" (For windows, inside the script the executable should be changed to the extension .exe).
3 - To check the output of the model, go to the folder simulation_results and open "stock_prediction_simulator_output_messages.txt".