In order to execute the code, it is necessary to previously install some packages. The python version should be 3.6.4
- numpy 1.14.5
- sklearn 0.19.1
- scipy 1.1.0
- pandas 0.23.1
- nltk 3.3
- keras 2.2.0
- tensorflow 1.8.0
- gensim 3.4.0
Alternatively you can execute the 'install_dependecies.sh' file which installs all the required packages in the correct version, but make sure all the packages are previously uninstalled in the machine.
To execute the code, just open a terminal inside the root directory of the project and execute the python file 'fake_trump.py', including (if wanted), some of the following flags.
--pop (population size)
--it (number of iterations)
--cross (crossover rate)
--mut (mutation rate)
--top (topics, separated by spaces)
Execution example:
python fake_trump.py --pop 200 --it 10 --cross 0.3 --mut 0.1 --top America news