Skip to main content

Markov Meme

PyPI version GitHub actions status

Wouldn't it be great to generate themed memes with Markov Models? Or just randomly generated text? I think so too.

Usage

Install

You can install from pypi

pip install markovmeme

or install from the repository directly:

$ git clone https://github.com/vsoch/markovmeme
$ python setup.py install

Usage

You can see the basic usage as follows. Basically, there is a "generate" command:

$ markov-meme 
usage: markov-meme [-h] {generate} ...

Markov Meme Generator

optional arguments:
  -h, --help  show this help message and exit

actions:
  actions for Markov Meme Generator

  {generate}  markovmeme actions
    generate  generate a meme

You can see the corpus available by looking at the generate help:

$ markov-meme generate --help
usage: markov-meme generate [-h] [--outfile OUTFILE] [--fontsize FONTSIZE]
                            [--font {OpenSans-Regular,Pacifico-Regular,Anton-Regular}]
                            [--corpus {hamlet,dr_seuss,trump_speech,the_office,office/stanley,office/toby,office/meredith,office/holly,office/creed,office/oscar,office/deangelo,office/david,office/dwight,office/kelly,office/phyllis,office/jim,office/nellie,office/gabe,office/clark,office/roy,office/karen,office/michael,office/andy,office/charles,office/kevin,office/pam,office/ryan,office/erin,office/robert,office/darryl,office/pete,office/jan,office/jo,office/angela}]
                            [--custom-corpus CUSTOM_CORPUS]
                            [--image CUSTOM_IMAGE] [--no-model]

optional arguments:
  -h, --help            show this help message and exit
  --outfile OUTFILE     the output file to save the image (defaults to
                        randomly generated png)
  --fontsize FONTSIZE   font size of text (if desired) defaults to 16
  --font {OpenSans-Regular,Pacifico-Regular,Anton-Regular}
                        choice of font (defaults to open sans)
  --corpus {hamlet,dr_seuss,trump_speech,the_office,office/stanley,office/toby,office/meredith,office/holly,office/creed,office/oscar,office/deangelo,office/david,office/dwight,office/kelly,office/phyllis,office/jim,office/nellie,office/gabe,office/clark,office/roy,office/karen,office/michael,office/andy,office/charles,office/kevin,office/pam,office/ryan,office/erin,office/robert,office/darryl,office/pete,office/jan,office/jo,office/angela}
                        the corpus to use to generate the meme, matches to
                        images.
  --custom-corpus CUSTOM_CORPUS
                        A custom corpus file, full path
  --image CUSTOM_IMAGE  A custom image file, full path
  --no-model            Don't generate a sentence from corpus, just randomly
                        select sentence.

Random Generation

To generate a meme from a random corpus, you can just run:

$ markov-meme generate

If the randomly selected corpus doesn't have matching images, you'll see

No images exist for corpus office/clark. Please specify --image.

Corpus and Image Selection

And you can run the same command again, or specify a specific corpus and image.

$ markov-meme generate --corpus office/michael --image markovmeme/data/images/office/michael1.png

The same can be done for specifying a custom corpus, which should be a text file with lines to generate the model from.

$ markov-meme generate --custom-corpus markovmeme/data/corpus/office/michael.txt --image markovmeme/data/images/office/michael2.png

Skip Model

If you don't want to use a model (and want real lines from some corpus) add --no-model

$ markov-meme generate --no-model

Interactive Python

For the most part, text size and number of characters are limited to fit reasonable within 1-3 lines. If you want to customize or otherwise play around with this, you can interact with the class directly:

from markovmeme.main import MemeImage
from markovmeme.text import generate_text

# corpus is absolute path, or relative to module's data/corpus folder
corpus = "office/michael"

text = generate_text(corpus=corpus, use_model=True, size=10)

# Set image to full path, or None to select based on corpus
meme = MemeImage(image=None, corpus=corpus)

# Add text generated, centered on top
meme.write_text(text, fontsize=args.fontsize, font=font)

# Leave outfile as None to generate random name
meme.save_image(None)

Support

Do you have a question? Or want to suggest a feature to make it better? Please open an issue!

Release files for markovmeme 0.0.1

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

Source distribution (sdist)

Source distribution for markovmeme 0.0.1
File Size Uploaded
markovmeme-0.0.1.tar.gz 51.8 MB Details

Built distribution (wheel)

Table of built distributions (wheels) for markovmeme 0.0.1
File Interpreter ABI Platform
markovmeme-0.0.1-py3.7.egg Legacy Egg format - - Details

Total release size: 103.8 MB

Release files / markovmeme-0.0.1.tar.gz

Download URL markovmeme-0.0.1.tar.gz
Size 51.8 MB
Tags Source
SHA-256 checksum
How to use checksums
bd46d74dcda01f73d978a91de696b5a1a077bda9263464188ac902333e1a9752
BLAKE2b-256 checksum
How to use checksums
aecabf345eb16403330c63d66890e5814219647d95b1f7df9cf1c9a39257880c
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/1.11.0 pkginfo/1.5.0.1 requests/2.21.0 setuptools/40.8.0 requests-toolbelt/0.9.1 tqdm/4.31.1 CPython/3.7.3

Release files / markovmeme-0.0.1-py3.7.egg

Download URL markovmeme-0.0.1-py3.7.egg
Size 51.9 MB
Tags Egg
SHA-256 checksum
How to use checksums
8518ffaf4a430efc31d3fd5e9d2d20ca3829e8e94785eefcaed8bb489e8fbe72
BLAKE2b-256 checksum
How to use checksums
87e1526c8d324573ce45da47d845e27929e9d5b66dd769e108f568038c77ede3
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/1.11.0 pkginfo/1.5.0.1 requests/2.21.0 setuptools/40.8.0 requests-toolbelt/0.9.1 tqdm/4.31.1 CPython/3.7.3

Release history Release notifications | RSS feed

This release

0.0.1 This release

2 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page