| ... | ... |
@@ -237,8 +237,8 @@ pca_biplot_plot<-setClass( |
| 237 | 237 |
params.points_to_label='entity', |
| 238 | 238 |
params.factor_name='entity', |
| 239 | 239 |
params.groups='entity', |
| 240 |
- params.sf='entity' |
|
| 241 |
- |
|
| 240 |
+ params.scale_factor='entity', |
|
| 241 |
+ params.style='entity' |
|
| 242 | 242 |
), |
| 243 | 243 |
prototype = list(name='Feature boxplot', |
| 244 | 244 |
description='plots a boxplot of a chosen feature for each group of a dataset.', |
| ... | ... |
@@ -268,6 +268,11 @@ pca_biplot_plot<-setClass( |
| 268 | 268 |
value=0.95, |
| 269 | 269 |
type='numeric', |
| 270 | 270 |
description='Scaling factor to apply to loadings. Default = 0.95.' |
| 271 |
+ ), |
|
| 272 |
+ params.style=entity(name='Plot style', |
|
| 273 |
+ value='points', |
|
| 274 |
+ type='character', |
|
| 275 |
+ description='Named plot styles for the biplot. [points]' |
|
| 271 | 276 |
) |
| 272 | 277 |
) |
| 273 | 278 |
|
| ... | ... |
@@ -307,7 +312,7 @@ setMethod(f="chart.plot", |
| 307 | 312 |
|
| 308 | 313 |
# plot |
| 309 | 314 |
A=data.frame("x"=P[,opt$components[1]]*sf*0.8,"y"=P[,opt$components[2]]*sf*0.8)
|
| 310 |
- SP=pca_scores_plot() |
|
| 315 |
+ out=pca_scores_plot() |
|
| 311 | 316 |
|
| 312 | 317 |
|
| 313 | 318 |
if (opt$style=='points') |
| 0 | 5 |
new file mode 100644 |
| ... | ... |
@@ -0,0 +1,40 @@ |
| 1 |
+# test PCA |
|
| 2 |
+test_that('PCA gives expected sum of sqaures for Iris data',{
|
|
| 3 |
+ # dataset |
|
| 4 |
+ D=iris_dataset() |
|
| 5 |
+ # PCA model |
|
| 6 |
+ M=PCA() |
|
| 7 |
+ # train the model |
|
| 8 |
+ M=model.train(M,D) |
|
| 9 |
+ # apply the model |
|
| 10 |
+ M=model.predict(M,D) |
|
| 11 |
+ # check the sum of squares |
|
| 12 |
+ expect_equal(M$ssx,9539.29) |
|
| 13 |
+ |
|
| 14 |
+ # using 1 component |
|
| 15 |
+ M$number_components=1 |
|
| 16 |
+ # train the model |
|
| 17 |
+ M=model.train(M,D) |
|
| 18 |
+ # apply the model |
|
| 19 |
+ M=model.predict(M,D) |
|
| 20 |
+ # check the sum of squares |
|
| 21 |
+ expect_equal(M$ssx,9539.29) |
|
| 22 |
+}) |
|
| 23 |
+ |
|
| 24 |
+test_that('PCA scores chart returns ggplot object',{
|
|
| 25 |
+ # dataset |
|
| 26 |
+ D=iris_dataset() |
|
| 27 |
+ # PCA model |
|
| 28 |
+ M=PCA() |
|
| 29 |
+ # train the model |
|
| 30 |
+ M=model.train(M,D) |
|
| 31 |
+ # apply the model |
|
| 32 |
+ M=model.predict(M,D) |
|
| 33 |
+ # chart |
|
| 34 |
+ C=pca_scores_plot(groups=D$sample_meta$Species) |
|
| 35 |
+ # plot |
|
| 36 |
+ gg=chart.plot(C,M) |
|
| 37 |
+ plot(gg) |
|
| 38 |
+ expect_true(is(gg,'ggplot')) |
|
| 39 |
+}) |
|
| 40 |
+ |
