Skip to content
Permalink
Browse files

updated README and DESCRIPTION for #20

  • Loading branch information...
pbiecek committed Aug 24, 2019
1 parent 777a717 commit a6e3aabb1618b8dce871e9ef3d48bf8a18ed1d47
Showing with 1,649 additions and 1,614 deletions.
  1. +33 −33 DESCRIPTION
  2. +14 −14 NAMESPACE
  3. +103 −102 README.md
  4. +299 −282 docs/articles/vignette_modelStudio.html
  5. +175 −175 docs/index.html
  6. +6 −6 docs/pkgdown.yml
  7. +320 −320 docs/reference/modelStudio.html
  8. +259 −259 docs/reference/modelStudioOptions.html
  9. +117 −117 man/modelStudio.Rd
  10. +85 −85 man/modelStudioOptions.Rd
  11. +238 −221 vignettes/vignette_modelStudio.Rmd
@@ -1,33 +1,33 @@
Package: dime
Title: Deep Interactive Model Explanations
Version: 0.1.6
Authors@R:
c(person("Hubert", "Baniecki", role = c("aut", "cre"), email = "hbaniecki@gmail.com"),
person("Przemyslaw", "Biecek", role = c("aut")))
Description: Automate explaining machine learning predictive models.
This package generates advanced interactive and animated model explanations in the form
of serverless HTML site. It combines R with D3.js to produce plots and descriptions
for local and global explanations. The whole is greater than the sum of its parts,
so it also supports EDA on top of that. ModelStudio is a fast and condensed way to get
all the answers without much effort. Break down your model and look into its ingredients
with only a few lines of code.
Depends: R (>= 3.5.0)
License: GPL
Encoding: UTF-8
LazyData: true
RoxygenNote: 6.1.1
Imports:
iBreakDown,
ingredients,
r2d3,
jsonlite
Suggests:
DALEX,
parallelMap,
randomForest,
knitr,
rmarkdown,
testthat
VignetteBuilder: knitr
URL: https://github.com/ModelOriented/dime
BugReports: https://github.com/ModelOriented/dime/issues
Package: dime
Title: Deep Interactive Model Explanations
Version: 0.1.6
Authors@R:
c(person("Hubert", "Baniecki", role = c("aut", "cre"), email = "hbaniecki@gmail.com"),
person("Przemyslaw", "Biecek", role = c("aut")))
Description: Automate explaining machine learning predictive models.
This package generates advanced interactive and animated model explanations in the form
of serverless HTML site. It combines R with D3.js to produce plots and descriptions
for local and global explanations. The whole is greater than the sum of its parts,
so it also supports EDA on top of that. ModelStudio is a fast and condensed way to get
all the answers without much effort. Break down your model and look into its ingredients
with only a few lines of code.
Depends: R (>= 3.5.0)
License: GPL
Encoding: UTF-8
LazyData: true
RoxygenNote: 6.1.1
Imports:
iBreakDown,
ingredients,
r2d3,
jsonlite
Suggests:
DALEX,
parallelMap,
randomForest,
knitr,
rmarkdown,
testthat
VignetteBuilder: knitr
URL: https://ModelOriented.github.io/dime/, https://github.com/ModelOriented/dime
BugReports: https://github.com/ModelOriented/dime/issues
@@ -1,14 +1,14 @@
# Generated by roxygen2: do not edit by hand

S3method(modelStudio,default)
S3method(modelStudio,explainer)
export(modelStudio)
export(modelStudioOptions)
importFrom(grDevices,nclass.Sturges)
importFrom(stats,aggregate)
importFrom(stats,predict)
importFrom(utils,head)
importFrom(utils,installed.packages)
importFrom(utils,setTxtProgressBar)
importFrom(utils,tail)
importFrom(utils,txtProgressBar)
# Generated by roxygen2: do not edit by hand
S3method(modelStudio,default)
S3method(modelStudio,explainer)
export(modelStudio)
export(modelStudioOptions)
importFrom(grDevices,nclass.Sturges)
importFrom(stats,aggregate)
importFrom(stats,predict)
importFrom(utils,head)
importFrom(utils,installed.packages)
importFrom(utils,setTxtProgressBar)
importFrom(utils,tail)
importFrom(utils,txtProgressBar)
205 README.md
@@ -1,102 +1,103 @@
# dime: Deep Interactive Model Explanations
### Automate Explaining Machine Learning Predictive Models

[![Project Status: Active – The project has reached a stable, usable state and is being actively developed.](https://www.repostatus.org/badges/latest/active.svg)](https://www.repostatus.org/#active)
[![Build Status](https://travis-ci.org/ModelOriented/dime.svg?branch=master)](https://travis-ci.org/ModelOriented/dime)
[![Coverage Status](https://codecov.io/gh/ModelOriented/dime/branch/master/graph/badge.svg)](https://codecov.io/github/ModelOriented/dime?branch=master)

This package generates advanced interactive and animated model explanations in the form
of serverless HTML site.

It combines **R** with **D3.js** to produce plots and descriptions
for local and global explanations. The whole is greater than the sum of its parts,
so it also supports EDA on top of that. ModelStudio is a fast and condensed way to get
all the answers without much effort. Break down your model and look into its ingredients
with only a few lines of code.

### [See an example](https://modeloriented.github.io/dime/demo.html)   [modelStudio - perks and features](https://modeloriented.github.io/dime/articles/vignette_modelStudio.html)

![](images/gif3.gif)

The package `dime` is a part of the [DrWhy.AI](http://drwhy.ai) universe.

Find more about model explanations in [Predictive Models: Visual Exploration, Explanation and Debugging](https://pbiecek.github.io/PM_VEE/) e-book.

------------------------------------------------------

## Installation

Install from GitHub:

```r
# dependencies
devtools::install_github("ModelOriented/ingredients")
devtools::install_github("ModelOriented/iBreakDown")
# dime
devtools::install_github("ModelOriented/dime")
```

Make sure that all dependencies are up-to-date with GitHub.

-------------------------------------------------------

## Demo

This package bases on `DALEX::explain()`.

```r
library("dime")
library("DALEX")
```

Create a model:

```r
titanic_small <- titanic_imputed[, c(1,2,3,6,7,9)]
titanic_small$survived <- titanic_small$survived == "yes"
model_titanic_glm <- glm(survived ~ gender + age + fare + class + sibsp,
data = titanic_small, family = "binomial")
```

Wrap it into an explainer:

```r
explain_titanic_glm <- explain(model_titanic_glm,
data = titanic_small[,-6],
y = titanic_small[,6],
label = "glm")
```

Pick some data points:

```r
new_observations <- titanic_small[1:4,]
rownames(new_observations) <- c("Lucas", "James", "Thomas", "Nancy")
```

Make a studio for the model:

```r
modelStudio(explain_titanic_glm, new_observations)
```

More examples [here](https://modeloriented.github.io/dime/articles/vignette_modelStudio.html).

![](images/gif4.gif)

------------------------------------------------------

## Save

You can save `modelStudio` using controls on the top of the RStudio Viewer
or with `r2d3::save_d3_html()` and `r2d3::save_d3_png()`.

![Save](images/controls.png)

------------------------------------------------------

## Cheat Sheet

![CheatSheet](images/basicCheatSheet.bmp)
# dime: Deep Interactive Model Explanations

[![Project Status: Active – The project has reached a stable, usable state and is being actively developed.](https://www.repostatus.org/badges/latest/active.svg)](https://www.repostatus.org/#active)
[![Build Status](https://travis-ci.org/ModelOriented/dime.svg?branch=master)](https://travis-ci.org/ModelOriented/dime)
[![Coverage Status](https://codecov.io/gh/ModelOriented/dime/branch/master/graph/badge.svg)](https://codecov.io/github/ModelOriented/dime?branch=master)

## Overview

The `dime` package automates explanation of Machine Learning predictive models This package generates advanced interactive and animated model explanations in the form
of serverless HTML site.

It combines **R** with **D3.js** to produce plots and descriptions
for local and global explanations. The whole is greater than the sum of its parts,
so it also supports EDA on top of that. ModelStudio is a fast and condensed way to get
all the answers without much effort. Break down your model and look into its ingredients
with only a few lines of code.

[See a demo](https://modeloriented.github.io/dime/demo.html) &emsp; [Read the vigniette: modelStudio - perks and features](https://modeloriented.github.io/dime/articles/vignette_modelStudio.html)

![](images/gif3.gif)

The package `dime` is a part of the [DrWhy.AI](http://drwhy.ai) universe.

## Installation

```{r}
# the easiest way to get ingredients is to install it from CRAN:
install.packages("ingredients")
# Install the development version from GitHub:
# dependencies
devtools::install_github("ModelOriented/ingredients")
devtools::install_github("ModelOriented/iBreakDown")
# dime
devtools::install_github("ModelOriented/dime")
```

## Demo

This package bases on DALEX explainers created with `DALEX::explain()`.

```r
library("dime")
library("DALEX")
```

Create a model:

```r
titanic_small <- titanic_imputed[, c(1,2,3,6,7,9)]
titanic_small$survived <- titanic_small$survived == "yes"
model_titanic_glm <- glm(survived ~ gender + age + fare + class + sibsp,
data = titanic_small, family = "binomial")
```

Wrap it into an explainer:

```r
explain_titanic_glm <- explain(model_titanic_glm,
data = titanic_small[,-6],
y = titanic_small[,6],
label = "glm")
```

Pick some data points:

```r
new_observations <- titanic_small[1:4,]
rownames(new_observations) <- c("Lucas", "James", "Thomas", "Nancy")
```

Make a studio for the model:

```r
modelStudio(explain_titanic_glm, new_observations)
```

More examples [here](https://modeloriented.github.io/dime/articles/vignette_modelStudio.html).

![](images/gif4.gif)

------------------------------------------------------

## Save

You can save `modelStudio` using controls on the top of the RStudio Viewer
or with `r2d3::save_d3_html()` and `r2d3::save_d3_png()`.

![Save](images/controls.png)

------------------------------------------------------

## Cheat Sheet

![CheatSheet](images/basicCheatSheet.bmp)



## Acknowledgments

Work on this package was financially supported by the 'NCN Opus grant 2016/21/B/ST6/02176'.

0 comments on commit a6e3aab

Please sign in to comment.
You can’t perform that action at this time.