scale_factor = 0.01
bound = 4
# First we define the callback function for the stream
def update(data):
global pos_x
global pos_y
data = data.channels_data
move = model(torch.from_numpy(data).float())
v = scale_factor * np.asarray(
[float(move[2] - move[0]),
float(move[1] - move[3])])
pos_x += v[0]
pos_y += v[1]
# truncate
pos_x = max(min(bound, pos_x), -1 * bound)
pos_y = max(min(bound, pos_y), -1 * bound)
# plot boundaries
ax.cla()
ax.plot([-1 * bound, bound, bound, -1 * bound, -1 * bound],
[-1 * bound, -1 * bound, bound, bound, -1 * bound])
# Draw circle at interpreted coordinates
ax.add_artist(plt.Circle((pos_x, pos_y), 0.1, color='blue'))
fig.canvas.draw()
# Start the stream
stream = StreamHandler(-1, update)
stream.start_stream()
# quit
if (user_input == "q"): break
# ping bci
if (user_input == "p"):
test = OpenBCICyton(daisy="True")
# record
elif (user_input == "r"):
row_to_modify = int(
input(
"Row to record to? [0 - Left] [1 - Up] [2 - Right] [3 - Down]")
)
frame_input = int(input("Recording how many frames?"))
mystream = StreamHandler(frame_input, collect_data)
mystream.start_stream()
print("Finished recording to", row_to_modify)
# save or save as
elif (user_input == "s" or user_input == "sa"):
if (LOADED and user_input == "s"):
np.save(LOADED_PATH, np.asarray(data))
else:
file_name = str(input("File path?"))
np.save(file_name, np.asarray(data))
# load
elif (user_input == "l"):
file_name = str(input("File path?"))
arrdata = np.load(file_name, allow_pickle=True)
for i in range(NUM_CONTROLS):
