0

My data were collected from the literature for a meta-analysis:

 Study      Eeff        Neff   Weight (%)
    1   34.83181476 50.04959144 0.1
    1   33.76824264 49.53210571 0.1
    1   32.67390287 47.77214774 0.1
    1   33.05520666 47.30254996 0.1
    2   33.19730252 44.21602149 0.3
    2   34.04709188 44.48002926 0.3
    3   33.22501606 38.96770954 0.3
    3   33.80766289 41.66974683 0.3
    3   33.06675158 42.66282841 0.3
    3   32.43069645 37.06958147 0.3
    4   34.72367611 41.31151977 0.25
    4   33.93254846 43.8526145  0.25
    4   33.22982756 44.99923374 0.25
    4   32.85979541 46.35648939 0.25
    5   32.85979541 43.53737788 0.35
    5   32.85979541 44.16612955 0.35
    5   32.85979541 40.10215682 0.35
    5   32.85979541 39.05910344 0.35

I have Eeff and Neff were obtained from 5 studies. First I made a scatter plot between Eeff and Neff normally. However, since data came from different studies where different sample size, experimental conditions...I weighted study by "inverse variance", i.e. the larger the variance the lesser the weight. Total weight is 100% (or 1). I want to make a new scatter plot where this "weight" could be included? Could someone help me please?

Thanks!

hn.phuong
  • 835
  • 6
  • 15
  • 24

1 Answers1

2

Using ggplot2, could do something like:

require('ggplot2')
aFile = read.table('yourdata.txt',header = T)     
p=ggplot(aFile,aes(Eeff,Neff))
p + geom_point(aes(colour = Weight,size = Weight))

Obviously you will want to play with colors and marker size, etc....

enter image description here

AGS
  • 14,288
  • 5
  • 52
  • 67