Skip to main content

Stock your experiments

Project description


**expstock** is a tool to manage results of experiments in machine learning, data analysis, simulation, etc.

When we try to integrate machine learning models or performe simulation using a mathematical model, we execute the same script or program many times with different parameters or logics.
In order to summalize or reproduce our experiments, it is necessary to take environmental information comprehensively.

expstock can automatically save environmental information with text files with adding simple implementation.

## Usage

There are two implementation types.
The simplest way is to surround the target mothod with the decorator as below.

from expstock import expstock

e = expstock.ExpStock(report=True, dbsave=True)
def run(a, b):
return a + b

e.append_param(a=a, b=b)
e.set_memo('This is the first experiment')
run(a, b)

But in case such as using Jupyter notebook, following implementation may be more convenient.

from expstock import expstock

e = expstock.ExpStock(dbsave=True)

e.append_param(a=a, b=b)
e.set_memo('This is the first experiment')
result = a + b
e.result = result

## Log Format

expstock saves environmental information with following directory structure. This is default setting, and we can change it.

├── <yyyymmdd_hhmmss>_<experiment_name>
│   ├── exec_time.txt
│   ├── git_diff.txt
│   ├── git_head.txt
│   ├── machine_info.txt
│   ├── memo.txt
│   ├── params.txt
│   ├── report.txt
│   ├── result.txt
│   ├── stderr.txt
│   └── stdout.txt


In the above, each file contains following information.

|file name | contents|user implementation|
|exec_time.txt |start time, finish time, execution time of the experiment| - |
|git_diff.txt | result of `git diff`| - |
|git_head.txt |result of `git log -n 1 --format=%H`| - |
|machine_info.txt |machine info such as os version and hostname which can get with `platform` which is python builtin package| - |
|memo.txt |memo for each experiments. | e = ExpStock(memo = 'hoge') or e.set_memo(hoge) |
|params.txt |experiment parameter. | e = ExpStock(params=[{'a': a}, {'b': b }]) or e.apend_params(a=a, b=b))|
|result.txt|return value of the experiment| e.result = func() or automatically set when using decorator|
|stdout.txt|result of sys.stdout| - |
|stderr.txt|result of sys.stderr| - |
|report.txt|summary of above information. if specify `report=True` when create Expstock instance, it creates report.txt| - |

In addition, we can save other files such as machine learning models in same directory with a simple command.

from sklearn.externals import joblib
joblib.dump(model, e.log_dirname)

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for expstock, version 0.1.3
Filename, size File type Python version Upload date Hashes
Filename, size expstock-0.1.3-py3-none-any.whl (13.4 kB) File type Wheel Python version py3 Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page