|
1 | 1 |
|
2 | 2 | # mlr3spatial
|
3 | 3 |
|
| 4 | +Package website: [release](https://mlr3spatial.mlr-org.com/) | |
| 5 | +[dev](https://mlr3spatial.mlr-org.com/dev/) |
| 6 | + |
4 | 7 | <!-- badges: start -->
|
| 8 | + |
5 | 9 | [](https://github.com/mlr-org/mlr3spatial/actions/workflows/r-cmd-check.yml)
|
6 | 10 | [](https://CRAN.R-project.org/package=mlr3spatial)
|
8 | 12 | [](https://stackoverflow.com/questions/tagged/mlr3)
|
9 | 13 | [](https://lmmisld-lmu-stats-slds.srv.mwn.de/mlr_invite/)
|
10 | 14 | <!-- badges: end -->
|
11 | 15 |
|
12 |
| -Package website: [release](https://mlr3spatial.mlr-org.com/) \| |
13 |
| -[dev](https://mlr3spatial.mlr-org.com/dev/) |
14 |
| - |
15 |
| -{mlr3spatial} is an extension package for spatial objects within the |
16 |
| -[mlr3](https://mlr3.mlr-org.com) ecosystem. |
17 |
| - |
18 |
| -## Feature Overview |
19 |
| - |
20 |
| -- Read training data from [sf](https://CRAN.R-project.org/package=sf) |
21 |
| - objects. |
22 |
| -- Predict on raster objects from the packages |
23 |
| - [{terra}](https://CRAN.R-project.org/package=terra), |
24 |
| - [{raster}](https://CRAN.R-project.org/package=raster) and |
25 |
| - [{stars}](https://CRAN.R-project.org/package=stars). |
26 |
| -- Write model predictions to raster files. |
27 |
| -- Predict large raster objects in parallel. |
28 |
| -- Read raster objects in chunks to avoid memory issues. |
29 |
| -- Built-in toy task |
30 |
| - [`leipzig`](https://mlr3spatial.mlr-org.com/dev/reference/leipzig.html). |
31 |
| - |
32 |
| -Check out |
33 |
| -[{mlr3spatiotempcv}](https://github.com/mlr-org/mlr3spatiotempcv) for |
| 16 | +*mlr3spatial* is the package for spatial objects within the |
| 17 | +[mlr3](https://mlr3.mlr-org.com) ecosystem. The package directly loads |
| 18 | +data from [sf](https://CRAN.R-project.org/package=sf) objects to train |
| 19 | +any mlr3 learner. The learner can predict on various raster formats |
| 20 | +([{terra}](https://CRAN.R-project.org/package=terra), |
| 21 | +[{raster}](https://CRAN.R-project.org/package=raster) and |
| 22 | +[{stars}](https://CRAN.R-project.org/package=stars)) and writes the |
| 23 | +prediction raster to disk. mlr3spatial reads large raster objects in |
| 24 | +chunks to avoid memory issues and predicts the chunks in parallel. Check |
| 25 | +out [mlr3spatiotempcv](https://github.com/mlr-org/mlr3spatiotempcv) for |
34 | 26 | spatiotemporal resampling within mlr3.
|
35 | 27 |
|
36 | 28 | ## Installation
|
|
76 | 68 | ## 3: 732737.2 5692469
|
77 | 69 | ## 4: 733169.3 5692777
|
78 | 70 | ## 5: 732202.2 5692644
|
79 |
| - ## --- |
| 71 | + ## --- |
80 | 72 | ## 93: 733018.7 5692342
|
81 | 73 | ## 94: 732551.4 5692887
|
82 | 74 | ## 95: 732520.4 5692589
|
@@ -118,20 +110,21 @@ plot(land_cover, col = c("#440154FF", "#443A83FF", "#31688EFF",
|
118 | 110 | ## FAQ
|
119 | 111 |
|
120 | 112 | <details>
|
121 |
| -<summary> |
122 |
| -Will mlr3spatial support spatial learners? |
123 |
| -</summary> |
124 |
| -<br> Eventually. It is not yet clear whether these would live in |
| 113 | + |
| 114 | +<summary>Will mlr3spatial support spatial learners?</summary> <br> |
| 115 | +Eventually. It is not yet clear whether these would live in |
125 | 116 | mlr3extralearners or in {mlr3spatial}. So far there are none yet.
|
| 117 | + |
126 | 118 | </details>
|
| 119 | + |
127 | 120 | <details>
|
128 |
| -<summary> |
129 |
| -Why are there two packages, {mlr3spatial} and {mlr3spatiotempcv}? |
130 |
| -</summary> |
131 |
| -<br> mlr3spatiotempcv is solely devoted to resampling techniques. There |
132 |
| -are quite a few and keeping packages small is one of the development |
133 |
| -philosophies of the mlr3 framework. Also back in the days when |
134 |
| -mlr3spatiotempcv was developed it was not yet clear how we want to |
135 |
| -structure additional spatial components such as prediction support for |
136 |
| -spatial classes and so on. |
| 121 | + |
| 122 | +<summary>Why are there two packages, {mlr3spatial} and |
| 123 | +{mlr3spatiotempcv}?</summary> <br> mlr3spatiotempcv is solely devoted to |
| 124 | +resampling techniques. There are quite a few and keeping packages small |
| 125 | +is one of the development philosophies of the mlr3 framework. Also back |
| 126 | +in the days when mlr3spatiotempcv was developed it was not yet clear how |
| 127 | +we want to structure additional spatial components such as prediction |
| 128 | +support for spatial classes and so on. |
| 129 | + |
137 | 130 | </details>
|
0 commit comments