zmsai 0.1.2

Ai solution for Luhmann's Zettelkasten

---

A command line utility for topic discovery and doc-linking within the Zettelkasten using AI approaches.

Installation

Install zmsai by executing the following command-

$ pip install zmsai

Test Run

Test run using dummy docs (see ./custom)

$ zmsai test

Usage

To learn n topics in your Zettelkasten at /path/to/your/zettelkasten/-

$ zmsai run -t n -p "/path/to/your/zettelkasten/"

This will create a metadata file meta.zms storing all the distributions exhibited by the documents in your Zettelkasten.

[Running the model] it may take a while. Hang tight!
[Data stored] ... kb

You can delete your metadata file by executing-

$ zmsai delete

These learnt distributions can be printed using zmsai display. You can pass an additional argument -w, the number of top most occuring words that you want to print from the distributions involing words.

To display doc-topic distribution-

$ zmsai display -d dt

To display topic-word distribution-

$ zmsai display -w n -d tw

To display doc-word distribution-

$ zmsai display -w n -d dw

To display vocabulary-

$ zmsai display -w n -d voc

To display all distributions at once-

$ zmsai display -w n -d all

or simply

$ zmsai display

This will take default value of 5 for nwords argument.

Manual

usage: zmsai [-h] [--path [PATH]] [--topics [TOPICS]] [--nwords [NWORDS]] [--distro [DISTRO]] [task]

positional arguments:
  task                  Provide task to perform [default : 'run'] [values : 'run', 'delete', 'display', 'man', 'test']

optional arguments:
  -h, --help            show this help message and exit
  --path [PATH], -p [PATH]
                        Provide directory of text files. [with : 'run'] [default : './custom']
  --topics [TOPICS], -t [TOPICS]
                        How many topics do you expect? [with : 'run'] [default : 'number of docs']
  --nwords [NWORDS], -w [NWORDS]
                        How many words per topic/doc do you want to display? [with : 'display'] [default : 5]
  --distro [DISTRO], -d [DISTRO]
                        What distributions do you want to display? [with : 'display'] [default : all] [values : 'dt', 'tw', 'dw', 'voc', 'all']

Dependency Graph

attrs==20.2.0
  - pytest==6.1.1 [requires: attrs>=17.4.0]
iniconfig==1.1.1
  - pytest==6.1.1 [requires: iniconfig]
joblib==0.17.0
  - scikit-learn==0.23.2 [requires: joblib>=0.11]
    - sklearn==0.0 [requires: scikit-learn]
numpy==1.19.2
  - scikit-learn==0.23.2 [requires: numpy>=1.13.3]
    - sklearn==0.0 [requires: scikit-learn]
  - scipy==1.5.3 [requires: numpy>=1.14.5]
    - scikit-learn==0.23.2 [requires: scipy>=0.19.1]
      - sklearn==0.0 [requires: scikit-learn]
pip==20.1.1
pluggy==0.13.1
  - pytest==6.1.1 [requires: pluggy>=0.12,<1.0]
py==1.9.0
  - pytest==6.1.1 [requires: py>=1.8.2]
pyparsing==2.4.7
  - packaging==20.4 [requires: pyparsing>=2.0.2]
    - pytest==6.1.1 [requires: packaging]
setuptools==46.4.0
six==1.15.0
  - packaging==20.4 [requires: six]
    - pytest==6.1.1 [requires: packaging]
threadpoolctl==2.1.0
  - scikit-learn==0.23.2 [requires: threadpoolctl>=2.0.0]
    - sklearn==0.0 [requires: scikit-learn]
toml==0.10.1
  - pytest==6.1.1 [requires: toml]
wheel==0.34.2

Contribution

Are welcome.

License

GNU General Public License v3 (GPLv3)