blesleep/sleepdata.py

from datetime import datetime
from os import path
import csv, time

import matplotlib.pyplot as plt
import matplotlib.animation as animation


sleep_data = { 
                'heartrate': {
                    'value_name': 'bpm',
                    'periods': [2, 5, 10, 15], 
                    'raw_data': [],
                    'averaged_data': [],
                    'last_hr': []
                    },
                'movement':{
                    'value_name': 'movement',
                    'periods': [10, 30, 60],
                    'raw_data': [],
                    'averaged_data': []
                    }
                } 
            

tick_seconds = 0.5
last_tick_time = None

datestamp = datetime.now().strftime("%Y_%m_%d")
csv_header_name_format = '{}_{}'
csv_filename_format = '{}_{}.csv'

plt.style.use('dark_background')
graph_figure = plt.figure()
graph_figure.canvas.set_window_title('blesleep')

graph_axes = graph_figure.add_subplot(1, 1, 1)
graph_data = {}

graph_displaytime_minutes = None

last_heartrate = 0

class Average_Gyro_Data():
    gyro_last_x = 0
    gyro_last_y = 0
    gyro_last_z = 0
    # Each gyro reading from miband4 comes over as a group of three,
    #     each containing x,y,z values.  This function summarizes the
    #     values into a single consolidated movement value.
    def process(self, gyro_data):
        gyro_movement = 0
        for gyro_datum in gyro_data:
            gyro_delta_x = abs(gyro_datum['gyro_raw_x'] - self.gyro_last_x)
            self.gyro_last_x = gyro_datum['gyro_raw_x']
            gyro_delta_y = abs(gyro_datum['gyro_raw_y'] - self.gyro_last_y)
            self.gyro_last_y = gyro_datum['gyro_raw_y']
            gyro_delta_z = abs(gyro_datum['gyro_raw_z'] - self.gyro_last_z)
            self.gyro_last_z = gyro_datum['gyro_raw_z']
            gyro_delta_sum = gyro_delta_x + gyro_delta_y + gyro_delta_z
            gyro_movement += gyro_delta_sum
        return gyro_movement


def write_csv(data, name):
    fieldnames = ['time']
    for fieldname in data[0]:
        if fieldname != 'time':
            fieldnames.append(fieldname)
            if name == 'raw':
                name = '{}_{}'.format(name, fieldname)

    csv_filename = csv_filename_format.format(datestamp, name)

    if not path.exists(csv_filename):
        open_handle = 'w'
    else:
        open_handle = 'a'

    with open(csv_filename, open_handle, newline='') as csvfile:
        csv_writer = csv.DictWriter(csvfile, fieldnames=fieldnames)
        if open_handle == 'w':
            csv_writer.writeheader() 
        if type(data) is list:
            for row in data:
                csv_writer.writerow(row)
        else:
            csv_writer.writerow(data)


def flush_old_raw_data(tick_time):
    for data_type in sleep_data:
        s_data = sleep_data[data_type]
        periods = s_data['periods']

        cleaned_raw_data = []
        old_raw_data = []
        
        for raw_datum in s_data['raw_data']:
            datum_age = tick_time - raw_datum['time']
            if datum_age < max(periods):
                cleaned_raw_data.append(raw_datum)
            else:
                old_raw_data.append(raw_datum)

        s_data['raw_data'] = cleaned_raw_data
        if old_raw_data:
            write_csv(old_raw_data, 'raw')


def flush_old_graph_data(graph_displaytime_minutes):
    graph_displaytime_seconds = graph_displaytime_minutes * 60
    tick_time = time.time()
    for data_type in sleep_data:
        s_data = sleep_data[data_type]
        cleaned_graph_data = []
        old_graph_data = []
        for avg_datum in s_data['averaged_data']:
            datum_age = tick_time - datetime.timestamp(avg_datum['time'])
            if datum_age < graph_displaytime_seconds:
                cleaned_graph_data.append(avg_datum)
            else:
                old_graph_data.append(avg_datum)
        s_data['averaged_data'] = cleaned_graph_data


def average_raw_data(tick_time):
    global last_heartrate
    timestamp = datetime.fromtimestamp(tick_time)
    csv_out = {'time': timestamp }

    for data_type in sleep_data:
        s_data = sleep_data[data_type]
        period_averages_dict = {'time': timestamp}
        periods = s_data['periods']
        value_name = s_data['value_name']

        flush_old_raw_data(tick_time)

        for period_seconds in periods:
            period_data = []
            period_averages_dict[period_seconds] = 0
            for raw_datum in s_data['raw_data']:
                datum_age_seconds = tick_time - raw_datum['time']
                if datum_age_seconds < period_seconds:
                    period_data.append(raw_datum[value_name])
                    
            if len(period_data) > 0:
                period_data_average = sum(period_data) / len(period_data)
            else:
                if data_type == "heartrate" and period_seconds == min(periods):
                    period_data_average = last_heartrate
                else:
                    period_data_average = 0

            period_averages_dict[period_seconds] = zero_to_nan(period_data_average)

            csv_header_field_name = csv_header_name_format.format(data_type, period_seconds)
            csv_out[csv_header_field_name] = zero_to_nan(period_data_average)

        s_data['averaged_data'].append(period_averages_dict)
    write_csv([csv_out], 'avg')


def process_gyro_data(gyro_data, tick_time):
    sleep_move = sleep_data['movement']
    value_name = sleep_move['value_name']
    gyro_movement = average_gyro_data.process(gyro_data)
    #print("Gyro: {}".format(gyro_movement))
    sleep_move['raw_data'].append({
        'time': tick_time,
        value_name: gyro_movement
    })


def process_heartrate_data(heartrate_data, tick_time):
    last_heartrate_count = 20
    print("BPM: " + str(heartrate_data))
    if heartrate_data > 0:
        value_name = sleep_data['heartrate']['value_name']
        sleep_data['heartrate']['raw_data'].append({
            'time': tick_time,
            value_name: heartrate_data
        } )

        if len(sleep_data['heartrate']['last_hr']) > last_heartrate_count:
            sleep_data['heartrate']['last_hr'].pop(0)
        sleep_data['heartrate']['last_hr'].append(heartrate_data)


def analyze_heartrate(hr_count):
    # Finds the pct change between the lowest HR in the last $hr_count samples and the current HR
    pct_heartrate_increase = 0
    if len(sleep_data['heartrate']['last_hr']) >= hr_count:
        last_heartrate_list = sleep_data['heartrate']['last_hr'][-hr_count:]
        last_heartrate_min = min(last_heartrate_list)
        current_heartrate = last_heartrate_list[-1]
        pct_heartrate_increase = int((current_heartrate - last_heartrate_min)/last_heartrate_min*100)
    return pct_heartrate_increase
    

def zero_to_nan(value):
    if value == 0:
        return (float('nan'))
    return int(value)


def update_graph_data():
    for data_type in sleep_data:
        s_data = sleep_data[data_type]
        
        avg_data = s_data['averaged_data']

        if len(avg_data) > 1:
            
            g_data = graph_data[data_type]
            data_periods = s_data['periods']

            starting_index = max([(len(g_data['time']) - 1), 0])
            ending_index = len(avg_data) - 1

            sleep_data_range = avg_data[starting_index:ending_index]

            for sleep_datum in sleep_data_range:
                g_data['time'].append(sleep_datum['time'])
                for period in data_periods:
                    if g_data['data'][period] != 'nan':
                        g_data['data'][period].append(sleep_datum[period])


def init_graph_data():
    for data_type in sleep_data:
        data_periods = sleep_data[data_type]['periods']
        graph_data[data_type] = {
            'time': [],
            'data': {}
        }
        for period in data_periods:
            graph_data[data_type]['data'][period] = []


def graph_animation(i):
    
    if len(graph_data) == 0:
        init_graph_data()

    flush_old_graph_data(graph_displaytime_minutes)
    update_graph_data()

    for data_type in graph_data:
        if len(graph_data[data_type]['time']) > 0:
            graph_axes.clear()
            break

    plotflag = False
    for data_type in sleep_data:
        s_data = sleep_data[data_type]
        g_data = graph_data[data_type]
        if len(g_data['time']) > 0:
            plotflag = True
            data_periods = sleep_data[data_type]['periods']
            for period in data_periods:
                axis_label = "{} {} sec".format(s_data['value_name'], period)
                graph_axes.plot(g_data['time'],
                                g_data['data'][period],
                                label=axis_label) 

    if plotflag:
        plt.legend()


def init_graph(graph_displaytime_mins=60, maximize=False):
    global graph_displaytime_minutes
    graph_displaytime_minutes = graph_displaytime_mins
    if maximize:
        figure_manager = plt.get_current_fig_manager()
        figure_manager.full_screen_toggle()
    
    ani = animation.FuncAnimation(graph_figure, graph_animation, interval=1000)
    plt.show()


if __name__ == 'sleepdata':
    average_gyro_data = Average_Gyro_Data()