New Feature, Real time graph for Amanuensis

Repository
https://github.com/to-the-sun/amanuensis

<p dir="auto"><br /> <a href="https://steemit.com/utopian-io/@to-the-sun/add-an-ever-updating-visual-graph-to-a-python-script-a-task-request-for-the-open-source-project-the-amanuensis-automated" target="_blank" rel="noreferrer noopener" title="This link will take you away from hive.blog" class="external_link">Link to the task request<span>This development contribution is submitted in association with task request by <a href="/@to-the-sun">@to-the-sun: <h3>What is Amanuensis? <p dir="auto"><span>The Amanuensis is an automated songwriting and recording system created by <a href="/@to-the-sun">@to-the-sun aimed at ridding the process of anything left-brained, so one need never leave a creative, spontaneous and improvisational state of mind, from the inception of the song until its final master. The program will construct a cohesive song structure, using the best of what you give it, looping around you and growing in real-time as you play. All you have to do is jam and fully written songs will flow out behind you wherever you go. <h3>New Feature <h5>The graph <p dir="auto">I implemented a graph for plotting the values in the dictionary <code>rhythm. Rhythm dictionary acted as a sparse array which gets updated as the time passes when the user starts recording. I had to plot the plot the data stored in the <code>rhythm dictionary from the moment referred in <code>spout[0]. The graph should also have to be updated in real time so I was needed to clear the plotted data after the plotting has been done. <strong>Matplotlib was used to plot the graph. <h3>Implementation <p dir="auto"><span>Python script: <a href="https://github.com/to-the-sun/amanuensis/blob/master/consciousness.py" target="_blank" rel="noreferrer noopener" title="This link will take you away from hive.blog" class="external_link">https://github.com/to-the-sun/amanuensis/blob/master/consciousness.py <p dir="auto">To make the graph real time I created a separate process using Python's <code>multiprocessing library. Also, a <code>queue was created to pass the data to be plotted to the <code>drawing_process in real time. The Queue was made non-blocking to prevent the graph window from freezing, waiting for new data. <h3>Relevant codes <p dir="auto">Funtion for graph drawing process <pre><code>def draw(queue): """ Process to draw & update the graph. """ # Initializing Graph xlim = stats['wake'] + 25 ylim = 2 plt.ion() figure = plt.figure() figure.subplots_adjust(bottom=0.22) figure.patch.set_facecolor('#eff3b9') figure.canvas.set_window_title('Rhythm') axes = figure.add_subplot(111) axes.set_facecolor('#eff3b9') axes.set_xlim([0, xlim]) axes.set_ylim([0, ylim]) axes.set_title('Rhythm') axes.set_xlabel('ms') axes.set_ylabel('cumulative beat intervals') try: x_data, y_data = queue.get(False) except que.Empty: x_data, y_data = new_data() l, = axes.plot(x_data, y_data, '#757981') manager = figure.canvas.manager manager.window.wm_geometry("+{}-{}".format(10, 20)) manager.resize(550, 200) figure.canvas.draw() figure.canvas.flush_events() # Update Loop while True: try: x_data, y_data = queue.get(False) ylim = np.amax(y_data) + 1 axes.set_ylim([0, ylim]) if stats['wake'] > xlim: xlim = stats['wake'] + 25 axes.set_xlim([0, xlim]) l.set_xdata(x_data) l.set_ydata(y_data) figure.canvas.draw() figure.canvas.flush_events() except que.Empty: try: figure.canvas.start_event_loop(0.5) except TclError: print("Graph closed!") break except TypeError: pass except TclError: break <p dir="auto">Function to prepare new data for plotting. <pre><code>def new_data(keys=None, values=None): """ Function to prepare data for plotting the graph. """ ydata = np.zeros(stats['wake'] + 25, dtype='int') xdata = np.linspace(0, stats['wake'] + 25, stats['wake'] + 25, dtype='int') if keys is not None: i = 0 j = 0 while j < len(values): try: if i in keys: ydata[i] = values[j] j += 1 i += 1 except Exception: break return xdata, ydata <p dir="auto">Removing plotted values from the <code>rhythm dictionary after plotting. <pre><code>def remove_past_moments(r_dict, spout): """ Removing keys from the rhythm dict which denotes moments in the past Returns new rhythm dict. """ for key in r_dict.copy(): if key - spout < 0: r_dict.pop(key) return r_dict <p dir="auto">Preparing new plot data and sending it via the queue <pre><code>try: # Stripping off past moments from the rhythm dict. rhythm = remove_past_moments(rhythm, spout[0]) lists = sorted(rhythm.items()) x, y = zip(*lists) x = np.array(x) - spout[0] x, y = new_data(x, y) except ValueError as v: pass except Exception as e: print(e) finally: try: q.put((x, y)) except Exception as e: print(e) <p dir="auto"><strong>A working implementation of the graph can be seen in this video: <a href="https://dlive.io/livestream/to-the-sun/8e75a760-8c98-11e8-9b20-0242ac110003" target="_blank" rel="noreferrer noopener" title="This link will take you away from hive.blog" class="external_link">Video Link <h4>GitHub Account <p dir="auto"><span><a href="https://github.com/ajmaln" target="_blank" rel="noreferrer noopener" title="This link will take you away from hive.blog" class="external_link">https://github.com/ajmaln