0

I am trying to create a SOM Hexagonal Map in R using the Kohonen Package.

I've encountered a couple of problems while running this package. Sometimes I am receiving a message that says:

"Error in sample.int(length(x), size, replace, prob) : cannot take a sample larger than the population when 'replace = FALSE'"

I receive this message although my sample is 1,000 observations and I use a 100x50 graph.

I am reproducing this code but haven't been very successful.

sebastian-montero
  • 379
  • 1
  • 4
  • 18
  • Are you using the same dataset as given in the repository? – Pragyaditya Das Mar 19 '17 at 05:47
  • Hi! The sample dataset is a subset of the original dataset. The original dataset is of ~10,000 observations while the sample is of ~1,000. Ideally I would like to map all of the observations in a heatmap. – sebastian-montero Mar 19 '17 at 05:57
  • I tried reproducing the same code, it gave me an error `cannot create vector of size 998 mb`. – Pragyaditya Das Mar 19 '17 at 06:33
  • Thanks! I updated the Question. I'm starting with this https://www.r-bloggers.com/self-organising-maps-for-customer-segmentation-using-r/ guide. The goal is to create https://github.com/geoss/som_visualization_r – sebastian-montero Mar 19 '17 at 06:47

0 Answers0