I've got a number of lists with pairs of columns of intersecting elements, here supposing sets 1, 2, 3, that can have a 1-to-1, 1-to-many and many-to-many relationship:
df1 = data.frame(
X1 = paste('set100',sample(10,8,replace=TRUE),sep=''),
X2 = paste('set200',sample(10,8,replace=TRUE),sep='')
)
df2 = data.frame(
X1 = paste('set100',sample(10,8,replace=TRUE),sep=''),
X3 = paste('set300',sample(10,8,replace=TRUE),sep='')
)
df3 = data.frame(
X2 = paste('set100',sample(10,8,replace=TRUE),sep=''),
X3 = paste('set300',sample(10,8,replace=TRUE),sep='')
)
And I want to create a merged matrix of the two-column lists to use them as input for eVenn. In eVenn's 4-way example, the lists are already in an object like this:
> head(res3)
liste_1_.194. liste_2_.149. liste_3_.366. Total_lists ratios
10345445 1 0 0 1 2.159987
10345762 1 1 0 2 2.223848
10345791 1 1 1 3 2.519503
10345824 0 0 0 0 NA
10346191 0 0 0 0 NA
10346843 0 0 0 0 NA
ratios ratios
10345445 NA NA
10345762 2.085687 2.264225
10345791 2.518024 2.668271
10345824 NA 36.246703
10346191 NA 2.527424
10346843 NA 3.852753
I want to be able to call the eVenn command on the merged data like this:
mergedmatrix <- create_the_merged_matrix_somehow(df1,df2,df3)
evenn(path_lists="test",res=mergedmatrix,ud=TRUE)
Any ideas?
Unfortunately your example is difficult to follow. I would suggest that you avoid using the eVenn package. The package documentation is very limited, and the examples are very cryptic. It does not produce a figure in the standard X11 (on unix systems) display, and will only print the results to file.
I would suggest you consider the limma package, available from Bioconductor. It has functionality to make nice Venn diagrams. You can see some examples here.
I guess that what you call create_the_merged_matrix_somehow() would be solved by vennCounts() in limma.