Metadata-Version: 2.1
Name: jupyter-cogram
Version: 0.7.1
Summary: Intuitive coding for Jupyter Notebook using natural language.
Home-page: https://cogram.ai
Author: ricwo
Author-email: r@cogram.ai
License: UNKNOWN
Keywords: machine learning,NLP,code generation,program synthesis
Platform: UNKNOWN
Classifier: Programming Language :: Python
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.6
Classifier: Programming Language :: Python :: 3.7
Classifier: Programming Language :: Python :: 3.8
Classifier: Development Status :: 4 - Beta
Classifier: Operating System :: MacOS
Classifier: Operating System :: Unix
Classifier: Operating System :: Microsoft :: Windows
Classifier: License :: Other/Proprietary License
Requires-Python: >=3.6
Description-Content-Type: text/markdown

<a href="https://cogram.ai">                                                        
<img src="https://uploads-ssl.webflow.com/61294dc1bd225d7c490b4389/61337287d439e010eae32f7f_logo_black_and_white_2048.png" 
width="250" align="right"/>
</a>

# Cogram: Intuitive coding with natural language

Cogram brings intuitive coding with natural language to Jupyter Notebook.

[![pypi Version](https://img.shields.io/pypi/v/jupyter-cogram.svg?style=flat-square&logo=pypi&logoColor=white)](https://pypi.org/project/jupyter-cogram/)
[![Cogram on Slack](https://img.shields.io/badge/Slack-pink.svg)](https://join.slack.com/t/slack-cng9493/shared_invite/zt-v79xffk1-oTkdfMJ9yLpd8E2YhU2l4Q)

## 📖 Documentation

| Documentation              |                                                                |
| -------------------------- | -------------------------------------------------------------- |
| 🚀️ **[Sign up]**        | Sign up to get your API token and get started!              |
| ⭐️ **[How to]**        | New to Cogram? Check out our videos on how to get started!              |
| 📚 **[Community]**      | Have questions our comments? Join our Slack!                             |

[sign up]: https://get.Cogram.ai
[how to]: https://youtu.be/8i1zTdQ1xKI
[community]: https://join.slack.com/t/slack-cng9493/shared_invite/zt-v79xffk1-oTkdfMJ9yLpd8E2YhU2l4Q

## Features

- AI-powered coding for Jupyter Notebooks
- Supports Python: ideal for data science tasks
- Cycle through different suggestions

## ⏳ Install cogram

### Requirements

- **API token**: If you don't have one yet, [sign up]
- **Operating system**: macOS · Linux · Windows
- **Python version**: Python 3.6+ (only 64 bit)
- **Package managers**: [pip]

[pip]: https://pypi.org/project/spacy/
[conda]: https://anaconda.org/conda-forge/spacy

### Installation

The easiest way to install Cogram for Jupyter Notebook is using pip.

```bash
pip install -U jupyter-cogram
jupyter nbextension enable jupyter-cogram/main
```

You can now start a new Jupyter Notebook server with 
```bash
jupyter notebook
```

and you're ready to go!


### Updating Cogram

The easiest way to upgrade to a new version of Cogram is using pip:

```bash
pip install -U jupyter-cogram
```

You'll then have to kill any active Jupyter Notebook servers and start a new one with 
```bash
jupyter notebook
```

## 📚 Use Cogram

### 🛫 First start

Once you've installed Cogram and open a Jupyter Notebook for the first time, you'll
 be asked for your API token (you can 
 find that under [My Account](https://get.cogram.ai/account)).
 
On first start, paste your API token into the prompt box. You can toggle the
 extension off and on by clicking the Cogram button 
 (<img align="center" width="14" src="https://uploads-ssl.webflow.com/61294dc1bd225d7c490b4389/6131d7249979f73249363dd0_icon_black_64.png" />) in the toolbar. The green
  circle indicates that Cogram is active. 
  
### 🔮 Prompting Cogram

You can prompt Cogram by writing a comment into a code cell starting with `##`. You
 finish off the prompt by typing `##` again. The status light will turn orange
 , indicating that Cogram is busy. 
 
For example, the prompt
```python
## fibonacci sequence ##
```
will generate different functions that produce the Fibonacci sequence.
 
 Once your code has been generated, you can 
 explore different options with the ← and → keys. If you're happy with a suggestion, 
 you can accept it by waiting for two seconds. Alternatively, you can accept it right
  away by hitting ⌘+⏎ (macOS) or Ctrl+⏎ (Windows & Linux).


