starhopp3r

GPIO Basics

First install Python's Raspberry Pi GPIO module, it should be installed by default but let's just be sure.

$ sudo apt-get install python-rpi.gpio

Then import the module into the python script.

starhopp3r

To solve this challenge I combined convolutional, long short-term memory and fully connected networks to form a CLDNN as described in Google's research paper. The neural network was implemented using TFLearn.

import gym
import random
import numpy as np
import tflearn
from tflearn.layers.core import input_data, dropout, fully_connected
from tflearn.layers.estimator import regression 

starhopp3r

import numpy as np
import os

# Features and labels
training_set_inputs = np.array([[0, 0, 1], [1, 1, 1], [1, 0, 1], [0, 1, 1]])
training_set_outputs = np.transpose(np.array([[0, 1, 1, 1]]))
# Predict
predict = np.array([[0, 0, 1]])
# Check if model has been saved
if not os.path.exists('model.npy'):

starhopp3r

import numpy as np
import os


class NeuralNetwork:
    def __init__(self):
        # Our training data
        self.X = np.array([[0, 0, 1], [1, 1, 1], [1, 0, 1], [0, 1, 1]])
        self.y = np.transpose(np.array([[0, 1, 1, 1]]))
        # Seed random number generator to produce the same

starhopp3r

import numpy as np
import os


class NeuralNetwork:
    def __init__(self):
        # Our training data
        self.X = np.array([[0, 0, 1], [1, 1, 1], [1, 0, 1], [0, 1, 1]])
        self.y = np.transpose(np.array([[0, 1, 1, 1]]))
        # Seed random number generator to produce the same

starhopp3r

import numpy as np
import os


class NeuralNetwork:
    def __init__(self):
        # Our training data
        self.X = np.array([[0, 0, 1], [1, 1, 1], [1, 0, 1], [0, 1, 1], [0, 0, 0], [0, 1, 0]])
        self.y = np.transpose(np.array([[0, 1, 1, 1, 1, 0]]))
        # Seed random number generator to produce the same

starhopp3r

from numpy import *

if __name__ == '__main__':
    run()

starhopp3r

from numpy import *

if __name__ == '__main__':
    run()

starhopp3r

def run():
    points = genfromtxt('data.csv', delimiter=',')
    # Hyperparameter
    learning_rate = 0.0001
    # Initial values: y = mx + c
    initial_c = 0
    initial_m = 0
    # Iterations
    num_iterations = 1000
    # Optimal values for m and c

starhopp3r

def gradient_descent_runner(points, starting_c, starting_m, learning_rate, num_iterations):
    c = starting_c
    m = starting_m
    # Iterate
    for i in range(num_iterations):
        c, m = step_gradient(c, m, array(points), learning_rate)
    return [c, m]