Exemple #1
0
import rippleTank as rt
import matplotlib.pyplot as plt

# creates a ripple tank
tank = rt.RippleTank()

# creates a source on the ripple tank
rt.Source(tank,
          rt.sineSource,
          xcorners=(-15, 15),
          ycorners=(10, 11),
          freq=10.0)

rt.Mask(tank).fromFunc(rt.singleSlit, ((-15, 15), (0, tank.dy)))

tank.simulateTime(2.0, animation_speed=0.5)

ani = tank.makeAnimation()
# ani.save('singleSlit.gif', writer='imagemagick')

plt.show()
Exemple #2
0
    Returns:
        np.ndarray: array with sine values on source positions.
    """
    t = source.rippletank.dt * i
    answer = np.zeros_like(source.X_grid)
    #value = np.sin(2*np.pi*source.freq*t + source.phase)
    # if value < 0:
    #     answer[source.positions] = value
    # else:
    #     # return None
    r = int(round(i % 10))
    answer[source.positions] = randF[r]
    #print(randF[r])
    return answer


# creates a source on the ripple tank
rt.Source(tank, rt.dropSource, freq=10)

tank.simulateTime(10)
frame = tank.captureFrame()
#data = tank.solvePoints(100)

#frame = data[99]

ani = tank.makeAnimation()
# ani.save('simpleSource.gif', writer='imagemagick')

plt.show()
Exemple #3
0
import rippleTank as rt
import matplotlib.pyplot as plt

# creates a ripple tank
tank = rt.RippleTank(bc='close')

# creates a source on the ripple tank
rt.Source(tank,
          rt.sineSource,
          xcorners=(-tank.dx, tank.dx),
          ycorners=(10 - tank.dy, 10 + tank.dy),
          freq=10.0)

width = (tank.dx**2 + tank.dy**2)**0.5
rt.Mask(tank).fromFunc(rt.halfCircleMask, (0, -4, 6, width, 'x', 'lower'))

tank.simulateTime(2.0, animation_speed=0.5)

ani = tank.makeAnimation()
# ani.save('halfCircle.gif', writer='imagemagick')

plt.show()
Exemple #4
0
import rippleTank as rt
import matplotlib.pyplot as plt

# creates a ripple tank
tank = rt.RippleTank()

x0, y0 = -5, 0
# creates a source on the ripple tank
rt.Source(tank,
          rt.sineSource,
          xcorners=(x0 - tank.dx, x0 + tank.dx),
          ycorners=(y0 - tank.dy, y0 + tank.dy),
          freq=10.0)
x1, y1 = 5, 0
rt.Source(tank,
          rt.sineSource,
          xcorners=(x1 - tank.dx, x1 + tank.dx),
          ycorners=(y1 - tank.dy, y1 + tank.dy),
          freq=5.0)

tank.simulateTime(2.0, animation_speed=0.5)

ani = tank.makeAnimation()
# ani.save('multipleSources.gif', writer='imagemagick')

plt.show()
Exemple #5
0
def evalOneMax(individual):
    tank = rt.RippleTank(xdim=(-(tank_width / 2), (tank_width / 2)),
                         ydim=(-(tank_height / 2), (tank_height / 2)),
                         deep=100.0)
    x = 0.0
    y = 0.0
    freq = 1
    phase = 1
    amplitude = 1
    global numEvals
    numEvals += 1

    def createSineSource(data):
        def newFunc(source, i):
            answer = np.zeros_like(source.X_grid)
            r = int(round(i % genomeSize))
            global prevSourceVal
            newVal = prevSourceVal + data[r]
            newVal = (1.0 * newVal) / (np.abs(newVal) + 0.5)
            #newVal = np.clip(newVal, -2.0, 2.0)
            answer[source.positions] = newVal
            prevSourceVal = newVal
            return answer

        return newFunc

    global prevSourceVal
    prevSourceVal = 0
    rt.Source(tank,
              createSineSource(individual),
              xcorners=(x - tank.dx - 300, x + tank.dx - 300),
              ycorners=(y - tank.dy, y + tank.dy),
              freq=freq,
              phase=phase,
              amplitude=amplitude)
    #rt.Source(tank, createSineSource(individual), xcorners = (x-tank.dx + 300, x+tank.dx + 300), ycorners=(y-tank.dy, y+tank.dy), freq=freq, phase = phase, amplitude = amplitude)

    data = tank.solvePoints(genomeSize)
    #with CodeTimer('create ripple'):
    #data = tank.solvePoints(2)

    # For each frame calculate the gradient
    fitness = 0
    successRays = 0
    start_index = genomeSize - 2
    num_items = genomeSize - start_index
    start_ray_test = 50

    for i in range(start_index, genomeSize - 1):
        #for i in range(0, 1):
        frame = data[i]

        #x, y = np.meshgrid(range(frame.shape[0]), range(frame.shape[1]))

        # fig = plt.figure()
        # ax = fig.add_subplot(111, projection='3d')
        # ax.plot_surface(x, y, frame)
        # plt.title('z as 3d height map')
        # plt.show()

        # show hight map in 2d
        # plt.figure()
        # plt.title('z as 2d heat map')
        # p = plt.imshow(frame)
        # plt.colorbar(p)
        # plt.show()

        gradients = np.gradient(frame)
        x_grad = gradients[1]
        y_grad = gradients[0]

        #x_grad[26][0] = -2

        #print("gradient: " + str(x_grad[26][0]))
        #print("arctan gradient: " + str(np.arctan(x_grad[26][0])))
        #print("sin arctan gradient: " + str(np.sin(np.arctan(x_grad[26][0]))))
        #print("arcsin refr arctan gradient: " + str(np.arcsin((1 / 1.333) * np.sin(np.arctan(x_grad[26][0])))))
        #print("90 degrees - arcsin refr arctan gradient: " + str((90.0 * np.pi / 180.0) - np.arcsin((1 / 1.333) * np.sin(np.arctan(x_grad[26][0])))))

        # x_reflected_gradients = np.tan((90.0 * np.pi / 180.0) - np.arcsin((1 / 1.333) * np.sin(np.arctan(x_grad)))) * -1
        # y_reflected_gradients = np.tan((90.0 * np.pi / 180.0) - np.arcsin((1 / 1.333) * np.sin(np.arctan(y_grad)))) * -1

        x_grad = np.diff(frame[49]) / units
        x_grad[3] = 3
        #x_grad = np.ones(99)
        x_reflected_gradients = np.tan(
            ((90.0 * np.pi / 180.0) - np.arctan(x_grad)) +
            (np.arcsin((1.0 / 1.333) * np.sin(np.arctan(x_grad))))) * -1.0
        #x_reflected_gradients = np.tan(np.ones(99) * (87.0 * np.pi / 180.0))

        test = np.tan((90.0 * np.pi / 180.0) -
                      (np.arcsin((1 / 1.333) * np.sin(np.arctan(3)))) +
                      np.arctan(3)) * 1.0

        a = np.arctan(3)
        b = (1 / 1.333) * np.sin(np.arctan(3))
        c = (np.arcsin((1 / 1.333) * np.sin(np.arctan(3))))

        x_distance = (np.arange(100) + np.array(np.ones(100))[:, None] -
                      51.0) * 1.0
        y_distance = np.transpose(x_distance) * 1.0

        #x_reflected_gradients = individual[1:]

        z_from_x_distance = x_distance[49][
            1:] * x_reflected_gradients * units * -1.0
        #z_from_y_distance = y_distance * y_reflected_gradients
        z_from_y_distance = 0

        # OLD CODE
        # # Where the gradient is small the normal is going to shoot to infinity, try caping it:
        # x_grad = np.where((x_grad < 0.0001) & (x_grad >= 0), 0.0001, x_grad)
        # x_grad = np.where((x_grad > -0.0001) & (x_grad <= 0), -0.0001, x_grad)
        # y_grad = np.where((y_grad < 0.0001) & (y_grad >= 0), 0.0001, y_grad)
        # y_grad = np.where((y_grad > -0.0001) & (y_grad <= 0), -0.0001, y_grad)
        # z_from_x_distance = x_distance * (1.0 / x_grad) * -1.0
        # z_from_y_distance = y_distance * (1.0 / y_grad) * -1.0

        z_intercept = z_from_x_distance + z_from_y_distance

        z_intercept = np.where(z_intercept > 0.0, 1.0, z_intercept)

        focal_point_z = -800

        z_intercept_targeted = z_intercept - focal_point_z

        #z_intercept_targeted = np.where(z_intercept_targeted > 0.0, 0.0, z_intercept_targeted)
        #z_intercept_targeted = np.where(z_intercept_targeted < -1000.0, -1000.0, z_intercept_targeted)
        z_intercept_targeted = (z_intercept_targeted) / (
            np.abs(z_intercept_targeted) + 0.5)

        fitness += np.sum(np.abs(
            np.square(z_intercept_targeted))) / (1.0 * float(num_items))

        target_size = 30.0
        z_intercept_point = np.where(
            (z_intercept_targeted > -(target_size / 2.0)) &
            (z_intercept_targeted < (target_size / 2.0)), 1.0, 0.0)

        #if i > start_ray_test:
        successRays += np.sum(z_intercept_point) / (
            50.0 * 1.0)  #float(num_items - start_ray_test))

        #fitness += np.sum(z_intercept_point) / (50.0) #* float(num_items))

    #opp_over_adj =  np.divide(y_grad, x_grad)
    #angles = np.arctan(opp_over_adj)

    # For each gradient calculate the light hitting the collector

    #fitness = np.sum(np.absolute(magish))

    #diff1 = np.absolute(img1_norm - frame1)

    #fitness1 = abs(0 - np.sum(diff1))

    global globFitness
    global globSuccessRays
    if globFitness > fitness:
        globFitness = fitness
        #save_image('img_norm.png', img1_norm)
        #save_image('z_intercept.png', z_intercept)
        #save_image('z_intercept_targeted.png', 1 / z_intercept_targeted)
        #save_image('z_intercept_point.png', z_intercept_point)
        #save_image('x_grad.png', x_grad)
        #save_image('y_grad.png', y_grad)
        save_image('frame.png', frame - 1)

        #tank.simulateTime(10)
        #ani = tank.makeAnimation()
        #ani.save('simpleSource.gif', writer='imagemagick')

        # Draw image
        im = Image.new('RGBA', (6000, 6000), (0, 0, 0, 255))
        draw = ImageDraw.Draw(im)
        z1 = 0.0
        scale = 10.0
        y_translate = 3800
        #x_grad = x_grad / 6
        #x_reflected_gradients = x_reflected_gradients / 6

        for i in range(1, 99):

            x1 = i * units * scale
            x2 = (i + 1) * units * scale
            z2 = ((frame[49][i]) * scale) + y_translate

            draw.line((x1, z1, x2, z2), fill=(255, 255, 255))

            grad_z = x_grad[i - 1] * scale * units
            draw.line((x1, z1, x2, z1 + (grad_z)), fill=(255, 255, 255))

            #x_middle = 60*(i + 0.5)
            #y_middle = (prevZ + grad_z)/2

            ray_length = 45

            if x_reflected_gradients[i - 1] < 10000 and x_reflected_gradients[
                    i - 1] > -10000:  # and i == 49:

                grad_z_refl = x_reflected_gradients[i - 1] * (scale * units)

                # init_x = ((60*i) + (60*(i+1)))/2.0
                # init_y = (prevZ + prevZ + grad_z) / 2.0

                x1 = x1 + 20

                #draw.line((x1, z1, x2, z1 + grad_z_refl), fill=(0, 255, 0))

                draw.line((x1, z1, x1 - (units * scale) * ray_length,
                           (z1 - grad_z_refl * ray_length)),
                          fill=(0, 255, 0))
                draw.line((x1, z1, x1 + (units * scale) * ray_length,
                           (z1 + grad_z_refl * ray_length)),
                          fill=(0, 255, 0))

            z1 = z2

        # Draw collector
        draw.ellipse(((units * scale) * 50.5 - 10, focal_point_z * scale +
                      y_translate + 1000 - (target_size * scale / 2.0),
                      (units * scale) * 50.5 + 10, focal_point_z * scale +
                      y_translate + 1000 + (target_size * scale / 2.0)),
                     fill=(255, 255, 0))

        # draw the z intercepts
        for t in z_intercept * scale * 1.0 + y_translate + 1000:
            draw.ellipse(((units * scale) * 50.5 - 10, (t) - 10,
                          (units * scale) * 50.5 + 10, (t) + 10),
                         fill=128)

        draw.line((0, y_translate, (units * scale * numCells), y_translate),
                  fill=128)

        im = im.transpose(Image.FLIP_TOP_BOTTOM)
        #im.show()
        im.save("./out/test" + str(numEvals) + ".png")

        # Lets dump out a file to view in ray tracer!!!!
        it = map(
            lambda x: "{\"x\":" + str(x[0] * units) + ",\"y\":" + str(x[1]) +
            ",\"arc\":false},", enumerate(frame[49]))
        coords = ''.join(it)[:-1]

        first = "{\"version\":2,\"objs\":[{\"type\":\"parallel\",\"p2\":{\"type\":1,\"x\":600,\"y\":50,\"exist\":true},\"p1\":{\"type\":1,\"x\":0,\"y\":50,\"exist\":true},\"p\":0.5},{\"type\":\"refractor\",\"path\":[" + coords
        last = ",{\"y\":-800,\"x\":" + str(
            genomeSize * units
        ) + ",\"arc\":false},{\"y\":-800,\"x\":" + str(
            0
        ) + ",\"arc\":false}],\"notDone\":false,\"p\":1.5}],\"mode\":\"light\",\"rayDensity_light\":0.1,\"rayDensity_images\":1,\"observer\":null,\"origin\":{\"x\":542.9644690819766,\"y\":70.20970720433743},\"scale\":0.7289999999999995}"

        allt = first + last

        text_file = open("Output.txt", "w")
        text_file.write(allt)
        text_file.close()

    if globSuccessRays < successRays:
        globSuccessRays = successRays
        print("Success rays: " + str(successRays))

    return fitness,
Exemple #6
0
import rippleTank as rt
import matplotlib.pyplot as plt

# creates a ripple tank
tank = rt.RippleTank()

# creates a source on the ripple tank
rt.Source(tank, rt.sineSource, freq=10)

tank.simulateTime(2.0, animation_speed=0.5)

ani = tank.makeAnimation()
# ani.save('simpleSource.gif', writer='imagemagick')

plt.show()