https://github.com/polaris000/snippets

A collection of snippets and functions that I regularly use in my workflows as a data scientist. These snippets are utility functions that speed up my work, and focus on Pandas, Numpy and visualization libraries.
https://github.com/polaris000/snippets

ipython pandas python utilities visualization

Last synced: 24 days ago
JSON representation

A collection of snippets and functions that I regularly use in my workflows as a data scientist. These snippets are utility functions that speed up my work, and focus on Pandas, Numpy and visualization libraries.

• Host: GitHub
• URL: https://github.com/polaris000/snippets
• Owner: Polaris000
• License: mit
• Created: 2023-02-25T05:03:55.000Z (over 1 year ago)
• Default Branch: main
• Last Pushed: 2023-02-25T07:42:41.000Z (over 1 year ago)
• Last Synced: 2024-10-12T11:21:03.673Z (24 days ago)
• Topics: ipython, pandas, python, utilities, visualization
• Language: Jupyter Notebook
• Homepage:
• Size: 18.6 KB
• Stars: 1
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
## Snippets

### About

A collection of snippets and functions that I regularly use in my workflows as a data scientist. These snippets are utility functions that speed up my work, and focus on Pandas, Numpy and visualization libraries.



### Pandas

| Snippet  | Code | Description |

| ------------- | ------------- | ------------ |

| ![](https://img.shields.io/badge/NEW-success/?style=flat-square) Boosted value_counts | [Code](./pandas/boosted_value_counts.ipynb)| This function improves the `value_counts` function by outputing absolute and normalized counts simultaneously, for faster analysis. It also sets the default value of `dropna` to False, so any NaNs that exist easily spotted.|

| ![](https://img.shields.io/badge/NEW-success/?style=flat-square) Pandarallel Configuration| [Code](./pandas/pandarallel_config.ipynb)| Pandas isn't great when it comes to handling large amounts of data, mainly because it natively uses only a single core. Pandarallel is a very straightforward alternative to parallelize pandas code. |

### IPython + Miscellaneous

| Snippet  | Code | Description |

| ------------- | ------------- | ------------ |

| ![](https://img.shields.io/badge/NEW-success/?style=flat-square) Color print in jupyter notebooks | [Code](./ipython/color_print.ipynb)| Coloring specific values in your output can be an easy way to highlight important information. While there are packages like `termcolor` or `colorama`, I find that simply using ANSI color outputs works best. |

| ![](https://img.shields.io/badge/NEW-success/?style=flat-square) Progress bars in jupyter notebooks | [Code](./ipython/tqdm_config.ipynb)| `tqdm` is a package that lets you create progress bars. While it has notebook specific versions via `tqdm_notebook`, I find that directly using `tqdm` works just as well without the hassle of setting up `ipywidgets` and `IProgress`.|