In [44]:

    

import functools
import sys

tasks = [
    {
        "name": "fast",
        "interval": 0.4,
        # do this if nothing else is running
        "priority": 1,
        "duration": 0.01,
        "nice": 0.0
    },
    {
        "name": "medium",
        "interval": 0.8,
        # do this if slow is not running
        "priority": 2,
        "duration": 0.01,
        "nice": 0.0
    },
    {
        "name": "slow",
        "interval": 2,
        # always do this 
        "priority": 3,
        "duration": 0.5,
        "nice": 0.0
    }
]

T0 = 0
t = 0
def wait_time(task, elapsed):
    """how long to wait for the interval to occur again"""
    interval = task['interval']
    wait = interval - (elapsed % interval)
    # we have to wait a bit longer to avoid floating point issues
    wait += sys.float_info.epsilon
    return wait
def dead_line(task, elapsed):
    interval = task['interval']
    wait = interval - (elapsed % interval)
    wait += sys.float_info.epsilon
    dead = wait + task['duration']
    return dead
    

for i in range(20):
    elapsed = t - T0
    next_wait_time = functools.partial(wait_time, elapsed=elapsed)
    
    # select the next task based on finish first
    first_task = min(
        tasks, 
        key=lambda x: next_wait_time(x) + x['duration']
    )
    
    # how long will this run?
    first_stop = next_wait_time(first_task) + first_task['duration']
    
    # see if this blocks any other tasks
    overlapping_tasks = []
    for task in tasks:
        if next_wait_time(task) <= first_stop:
            overlapping_tasks.append(task)
    
    # select the task with the highest priority
    selected_task = max(overlapping_tasks, key=lambda x: x['priority'])
    # how long should we wait for it
    selected_wait = next_wait_time(selected_task) 
    
    # TODO: give bonus to missed tasks
    
    # wait for time to be executed
    t += selected_wait
    print('t: %.2f calling function %s' % (t, selected_task['name']))
    # execute it
    t += selected_task['duration']
    # update duration estimate


    

    




t: 0.40 calling function fast
t: 0.80 calling function medium
t: 1.20 calling function fast
t: 1.60 calling function medium
t: 2.00 calling function slow
t: 2.80 calling function fast
t: 3.20 calling function medium
t: 3.60 calling function fast
t: 4.00 calling function slow
t: 4.80 calling function medium
t: 5.20 calling function fast
t: 5.60 calling function medium
t: 6.00 calling function slow
t: 6.80 calling function fast
t: 7.20 calling function medium
t: 7.60 calling function fast
t: 8.00 calling function slow
t: 8.80 calling function medium
t: 9.20 calling function fast
t: 9.60 calling function medium

In [47]:

    

t = 0
for i in range(20):
    elapsed = t - T0
    next_wait_time = functools.partial(wait_time, elapsed=elapsed)
    
    next_dead_line = functools.partial(dead_line, elapsed=elapsed)
    
    # select the next task based on finish first
    first_task = min(
        tasks, 
        key=lambda x: next_dead_line(x)
    )
    
    # how long will this run?
    first_stop = next_dead_line(task)
    
    selected_task = first_task
    # TODO: give bonus to missed tasks
    
    
    # wait for time to be executed
    t += next_wait_time(selected_task)
    print('t: %.2f calling function %s' % (t, selected_task['name']))
    # execute it
    t += selected_task['duration']
    # update duration estimate


    

    




t: 0.40 calling function fast
t: 0.80 calling function fast
t: 1.20 calling function fast
t: 1.60 calling function fast
t: 2.00 calling function fast
t: 2.40 calling function fast
t: 2.80 calling function fast
t: 3.20 calling function fast
t: 3.60 calling function fast
t: 4.00 calling function fast
t: 4.40 calling function fast
t: 4.80 calling function fast
t: 5.20 calling function fast
t: 5.60 calling function fast
t: 6.00 calling function fast
t: 6.40 calling function fast
t: 6.80 calling function fast
t: 7.20 calling function fast
t: 7.60 calling function fast
t: 8.00 calling function fast

In [46]:

    

intervals = [task['interval'] for task in tasks]
min_interval = min(intervals)
fractions = [task['interval'] / min_interval for task in tasks]
fractions
import numpy as np
a = task['interval']/ min_interval
all([fraction.is_integer() for fraction in fractions])


    

    
Out[46]:





True

In [ ]: