Using clocks

Advanced timing

ticktock gives you different ways to specify which parts of your code you want to time: functions, a decorator, or a context manager.

tick and tock

Use tick to create a Clock and signal the start of a measurement:

from ticktock import tick

t = tick() # t is a Clock instance

Then, use the tock method to signal the end of the measurement for this clock:

t.tock()

Naming

By default, tick creates a clock named after where it is first created (e.g. path/to/the/code.py:line_number), and tock is named after the line in which it is called.

As a result, a Clock is typically renderer as path/to/the/code.py:start-stop where start and stop are the line numbers of the tick and tock.

You can name the beginning and end of timers:

clock = tick("beginning")
clock.tock("end")

Which will then be displayed as ⏱️ [beginning-end] 1ms count=1

Clock identity

tick and tock function by recording the specific line in your code that they are created at.

This allows ticktock to aggregate times together although the Clock is redefined everytime the code is visited:

for _ in range(100):
    t = tick() # <- this is always the same object
    ...
    t.tock()
    id(t) # <- this is always the same value

Context manager

It is possible to use ticktock as a context manager to track the timing of a chunk of code:

from ticktock import ticktock

with ticktock():
    time.sleep(1)

Naming

By default the name of the clock with a context manager is the filename and line numbers, but you can also provide a name to the context manager:

with ticktock(name="some name"):
    pass

Function decorator

ticktock doubles as a decorator that tracks the timing of each call to a function:

from ticktock import ticktock

@ticktock
def f():
    time.sleep(1)

f()

Naming

Naming ticktock function decorators By default the name of the clock with a function decorator is the name of the function:

@ticktock
def f():
    pass

Will render as: ⏱️ [f] 1ms count=1

You can also provide a name to the decorator

@ticktock(name="some name")
def f():
    pass

Will render as: ⏱️ [some name] 1ms count=1

Multiple Clocks

Of course, can create multiple independent ticks, which will appear as separate clocks:

for _ in range(1000):
    clock = tick()
    # do some work
    time.sleep(1)
    clock.tock()

    clock = tick()
    # do some other work
    time.sleep(0.5)
    clock.tock()

A clock can also have a multiple tocks, which will be displayed as different lines

t = tick("start")
time.sleep(1)
if k % 2 == 1:
    time.sleep(1)
    t.tock("one")
else:
    t.tock("two")

Which will lead to two lines being renderered:

⏱️ [start-one] 1s count=1
⏱️ [start-two] 2s count=1