https://github.com/machow/wrestlr

Why debate R vs python, when you can convert R to python?
https://github.com/machow/wrestlr

Last synced: 3 months ago
JSON representation

Why debate R vs python, when you can convert R to python?

• Host: GitHub
• URL: https://github.com/machow/wrestlr
• Owner: machow
• Created: 2020-04-23T16:49:22.000Z (over 4 years ago)
• Default Branch: master
• Last Pushed: 2020-04-24T14:41:47.000Z (over 4 years ago)
• Last Synced: 2024-04-26T20:46:58.120Z (8 months ago)
• Language: Python
• Homepage:
• Size: 145 KB
• Stars: 18
• Watchers: 5
• Forks: 1
• Open Issues: 1
• Metadata Files:
  • Readme: README.Rmd

Awesome Lists containing this project

README

        
# wrestlr

[![Binder](https://mybinder.org/badge_logo.svg)](https://mybinder.org/v2/gh/machow/wrestlr/master)

The language wars are over.

wrestlr uses python to do a two step dance:

* parse R code that uses [dplyr](https://github.com/tidyverse/dplyr).

* spit out python code that runs using the dplyr port [siuba](http://github.com/machow/siuba).

## Basic use

```{python}

import wrestlr

r_code = """

  mtcars %>%

    filter(hp < 200)

"""

wrestlr.rlang_convert(r_code)

```

It's much nicer to use IPython cell magic, though!

```{python}

import wrestlr

# %load_ext wrestlr

```

```{python}

# %%wrestlr --print

1

'a'

TRUE

NULL

x$y

x[["y"]]

```

Here is another example with ggplot.

```{python}

# %%wrestlr --print

mtcars %>%

  select(hp, mpg, cyl) %>%

  ggplot(aes(hp, mpg)) +

  geom_point() +

  facet_wrap(~cyl)

```

## Executing 

Instead of printing code, you can execute it using the `--execute` option. First, we'll import some python functions.

```{python}

# import wrestlr

# # %load_ext wrestlr

import pandas as pd

from siuba import _, mutate, group_by, ungroup

from siuba.data import mtcars

from plotnine import *

def factor(x):

    return pd.Categorical(x)

```

Next we convert and execute the code.

```{python}

# %%wrestlr --print --execute --black

mtcars %>%

  group_by(cyl) %>%

  mutate(demeaned_mpg = mpg - mean(mpg)) %>%

  ungroup() %>%

  ggplot(aes(factor(cyl), demeaned_mpg)) +

  geom_boxplot()

```

## Learning more

See these example notebooks

| name | binder | description |

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

| [gallery](docs/gallery.ipynb) | [![](https://mybinder.org/badge_logo.svg)][b-gallery] | Walk through rules wrestlr uses during conversion |

| [cell_magic](docs/cell_magic.ipynb) | [![](https://mybinder.org/badge_logo.svg)][b-cell-magic] | Get to know the %%wrestlr cell magic |

| [debugging.ipynb](docs/debugging.ipynb) | [![](https://mybinder.org/badge_logo.svg)][b-debugging] | Debugging the parser, AST, or siuba conversion | 

| [translate-tidytuesday](docs/translate-tidytuesday.ipynb) | [![](https://mybinder.org/badge_logo.svg)][b-translate-tidytuesday] | Translating and executing the first half of a tidy tuesday R analysis |

[b-gallery]: https://mybinder.org/v2/gh/machow/wrestlr/master?filepath=docs/gallery.ipynb

[b-cell-magic]: https://mybinder.org/v2/gh/machow/wrestlr/master?filepath=docs/cell_magic.ipynb

[b-debugging]: https://mybinder.org/v2/gh/machow/wrestlr/master?filepath=docs/debugging.ipynb

[b-translate-tidytuesday]: https://mybinder.org/v2/gh/machow/wrestlr/master?filepath=docs/translate-tidytuesday.ipynb