Indicates progress during runtime, while keeping track of time consumed by user-defined tasks. Not quite a progress bar, not quite a profiler, but something in between which I personally have found to be very handy, especially when working with computationally intensive programs.
Consider this simple dummy program:
from tasktimer import TaskTimer
from time import sleep
timer = TaskTimer()
for n in timer.range(40):
timer.task('Assembling load vector (b)')
sleep(0.1) # Dummy
timer.task('Solving linear system Au=b')
sleep(0.5) # Dummy
print(timer)
The task() method is used to indicate that from now on the program is working on another task. TaskTimer will indicate progress, and display the current task, for instance:
Step 11/40 (28%). 17.5s of 24.1s remaining. Assembling load vector (b)
When all tasks are completed, print(timer) can be used to print statistics of the tasks:
# Mean StDev Total % Assembling load vector (b) 40 101ms 89.2us 4.02s 17 Solving linear system Au=b 40 501ms 140us 20.0s 83 Total 24.1s 100
Many more options than demonstrated here are available, and the functionality is extensively documented in the docstrings. For instance, customization of the progress string (see TaskTimer), iterating over general iterables using the iterate() method, sorting the statistics (see the summary() method). The time tracking of the individual tasks are performed by the Timer class, which can also be used stand-alone.
Install from PyPI using pip:
pip install TaskTimer
Or download the GitHub repository https://github.com/sigvaldm/TaskTimer.git and run:
python setup.py install