flowshow

Just a super thin wrapper for Python tasks that form a flow.

Installation

uv pip install flowshow

Usage

Flowshow provides a @task decorator that helps you track and visualize the execution of your Python functions. Here's how to use it:

import time
import random

from flowshow import task

# Turns a function into a Task, which tracks a bunch of stuff
@task
def my_function(x):
    time.sleep(0.5)
    return x * 2

# Tasks can also be configured to handle retries
@task(retry_on=ValueError, retry_attempts=10)
def might_fail():
    time.sleep(0.5)
    if random.random() < 0.5:
        raise ValueError("oh no, error!")
    return "done"

@task
def main_job():
    print("This output will be captured by the task")
    for i in range(3):
        my_function(10)
        might_fail()
    return "done"

# Run like you might run a normal function
main_job()

Once you run your function you can expect some nice visuals, like this one:

main_job.plot()

You can also inspect the raw data yourself by running:

main_job.last_run.to_dict()

Show the full dictionary.

{
  "task_name": "main_job",
  "start_time": "2025-02-04T21:25:17.045576+00:00",
  "duration": 8.864794875029474,
  "inputs": {},
  "error": None,
  "retry_count": 0,
  "end_time": "2025-02-04T21:25:25.909997+00:00",
  "logs": "This output will be captured by the task\n",
  "output": "done",
  "subtasks": [
    {
      "task_name": "my_function",
      "start_time": "2025-02-04T21:25:17.045786+00:00",
      "duration": 0.5050525842234492,
      "inputs": {
        "arg0": 10
      },
      "error": None,
      "retry_count": 0,
      "end_time": "2025-02-04T21:25:17.550808+00:00",
      "logs": "",
      "output": 20
    },
    {
      "task_name": "might_fail",
      "start_time": "2025-02-04T21:25:17.550853+00:00",
      "duration": 0.5053939162753522,
      "inputs": {},
      "error": None,
      "retry_count": 0,
      "end_time": "2025-02-04T21:25:18.056233+00:00",
      "logs": "",
      "output": "done"
    },
    {
      "task_name": "my_function",
      "start_time": "2025-02-04T21:25:18.056244+00:00",
      "duration": 0.5052881669253111,
      "inputs": {
        "arg0": 10
      },
      "error": None,
      "retry_count": 0,
      "end_time": "2025-02-04T21:25:18.561502+00:00",
      "logs": "",
      "output": 20
    },
    {
      "task_name": "might_fail",
      "start_time": "2025-02-04T21:25:18.561516+00:00",
      "duration": 2.1351009169593453,
      "inputs": {},
      "error": None,
      "retry_count": 0,
      "end_time": "2025-02-04T21:25:20.696477+00:00",
      "logs": "",
      "output": "done"
    },
    {
      "task_name": "my_function",
      "start_time": "2025-02-04T21:25:20.696511+00:00",
      "duration": 0.5026454580947757,
      "inputs": {
        "arg0": 10
      },
      "error": None,
      "retry_count": 0,
      "end_time": "2025-02-04T21:25:21.199158+00:00",
      "logs": "",
      "output": 20
    },
    {
      "task_name": "might_fail",
      "start_time": "2025-02-04T21:25:21.199213+00:00",
      "duration": 4.711003000382334,
      "inputs": {},
      "error": None,
      "retry_count": 0,
      "end_time": "2025-02-04T21:25:25.909979+00:00",
      "logs": "",
      "output": "done"
    }
  ]
}

You can also get a flat representation of the same data in a dataframe via:

main_job.to_dataframe()

This is what it looks like in Marimo when you evaluate this. Note that we also track the logs of the print statements for later inspection.

Multiple runs

If you run the function multiple times you can also inspect multiple runs:

main_job.runs

This can be useful, but most of the times you're probably interested in the last run.

Release files for flowshow 0.2.2

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for flowshow 0.2.2
File	Size	Uploaded
flowshow-0.2.2.tar.gz	11.2 MB	Apr 30, 2025	Details

Built distribution (wheel)

Table of built distributions (wheels) for flowshow 0.2.2
File	Interpreter	ABI	Platform	Reset
flowshow-0.2.2-py3-none-any.whl 12.9 kB Apr 30, 2025	Python 3	none	any	Details

Total release size: 11.2 MB

Release files / flowshow-0.2.2.tar.gz

Download URL	flowshow-0.2.2.tar.gz
Size	11.2 MB
Tags	Source
SHA-256 checksum How to use checksums	`49492ef3758075909ed4e41dfe8450d6862b9a0149a874c819fe2bb1909ec9cf`
BLAKE2b-256 checksum How to use checksums	`46bf4c7d892768bc68f72c8a88d87ec1e6642b16f4055553fe56ec6ff0c9e112`
Upload date	Apr 30, 2025
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`uv/0.6.7`

Release files / flowshow-0.2.2-py3-none-any.whl

Download URL	flowshow-0.2.2-py3-none-any.whl
Size	12.9 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`017c5abc64b4b043514925215db91fdabdbe2e5e229152781236282a7fbc52f3`
BLAKE2b-256 checksum How to use checksums	`1486002823427c96500baf33fd40c33ffddeff91f3ae509129406cd4487665ac`
Upload date	Apr 30, 2025
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`uv/0.6.7`