adjclust icon indicating copy to clipboard operation
adjclust copied to clipboard

change plotSim with ggplot2

Open tuxette opened this issue 6 years ago • 1 comments

tuxette avatar Dec 10 '18 11:12 tuxette

work in progress from commit 9383f3ffa0f9175a0a56c1855a5efae6ab8359b1

  • [x] check the addition of ggplot2 elements like palette, name of the legend, ... (choose defaults that are the best)
  • [x] check if prior.count should be added
  • [x] extend to sparse matrices in a smart way
  • [x] add clustering
  • [x] add dendrograms
  • [x] add xaxis
  • [x] check for inputs (for xaxis, clustering)
  • [x] document
  • [x] check for dependencies

tuxette avatar Nov 08 '22 13:11 tuxette

Addressing check the addition of ggplot2 elements like palette, name of the legend, ... (choose defaults that are the best):

  • [X] palette : default viridis palette; custom palette can be added as in examples
  • [X] name of the legend : added the argument legendName with default "intensity"
  • [X] name of the plot : added the argument main with default to NULL (no title) added in 172ee05d8735d7ffcce3e067b157a0115fdc6c26

tuxette avatar Dec 08 '22 14:12 tuxette

Addressing check if prior.count should be added

Added a prior count in fea479fb32bf022125c4d102a34cb4bb166e0ffe (similar to edgeR::cpm).

tuxette avatar Dec 08 '22 15:12 tuxette

Addressing extend to sparse matrices in a smart way (the "smart way" might not be quite perfect though):

  • [x] dgCMatrix-class
  • [x] dsCMatrix-class
  • [x] dist
  • [x] HTCexp
  • [x] SnpMatrix

Note on Dec. 8th: work in progress on branch dev-ggplot2 (but it might be simply easier and more efficient to just convert sparse matrices do dense ones). Think of what is desired behavior for zeros... (maybe just skip zeros when sparse matrices are used and print a message for the user; in addition, it might be needed to set xlim and ylim and that case: they can be found from the matrix dimension; finally, we might think to add a grey triangle in the background to display the chromosome).

Note on Jan. 12th: done this way for dsCMatrix (grey background, only non negative values are plotted)

fixed in commit 5096e6866b65edbbce164214970c56cbf0de383f

tuxette avatar Dec 08 '22 15:12 tuxette

add xaxis fixed in commit 8b0a78c9c6cea34a066b334747618c68c314209e

tuxette avatar Feb 02 '23 10:02 tuxette

add clustering in commit 4fb14544dd32604f0177eb8960036ca0d4c8e587

tuxette avatar Feb 06 '23 18:02 tuxette

add dendrogram: a first version is provided in commit 906e1d4fbdbd02bfc5629a596524639258fd2be9 Next to do:

  • [x] check on iris data
  • [x] factorize code (create a function for coordinates transformation in dendrograms)
  • [x] make axis work with dendrograms
  • [x] make clustering work with dendrograms

tuxette avatar Mar 07 '23 15:03 tuxette

finished dendrogram in commit 06a406a3029f975bebb9d21dea86992342bafbf8

tuxette avatar Mar 16 '23 10:03 tuxette

added documentation in commit 570bc9d3f83d6d48ccc39153a65815681df83892

tuxette avatar Mar 16 '23 14:03 tuxette

fixed dependencies solved in commit b29178e88bd83c779a3150a0f32ccc5565129e02

tuxette avatar Mar 16 '23 16:03 tuxette