-1

My PCA result using prcomp() function is summarised and plot as followings. How to interpret the plot results? It shows in some online article that the points present the amount of variance attributed to the different principal components. However, the value seems not matching with any of the statistics, e.g., standard deviation, the proportion of variance, or cumulative proportion. 

> summary(data_pca)

enter image description here

> plot(data_pca,type="lines")

enter image description here

zx8754
  • 52,746
  • 12
  • 114
  • 209
HappyCoding
  • 5,029
  • 7
  • 31
  • 51
  • 1
    Since this is not a coding question, this seems more appropriate as a question over at [CrossValidated](https://stats.stackexchange.com/). On top of that, you can find **a lot** of information on the interpretation of these (commonly called) **scree plots**. Please take some time researching, and if still unclear I would suggest posting a question over at CV. – Maurits Evers Jun 14 '18 at 05:57
  • 1
    Calculate the squared standard deviations and compare with the plot. Also, read `help("screeplot")`. – Roland Jun 14 '18 at 05:59
  • @Roland, thanks both. I understood now. – HappyCoding Jun 14 '18 at 06:04

1 Answers1

0

I got the hint from @Roland and @Maurits. Here, the variance is exactly the square of standard deviation.

HappyCoding
  • 5,029
  • 7
  • 31
  • 51