OpenMendel / SnpArrays.jl

Compressed storage for SNP data
https://openmendel.github.io/SnpArrays.jl/latest
Other
44 stars 9 forks source link

no A_mul_B!, Ac_mul_B! routines #2

Open klkeys opened 8 years ago

klkeys commented 8 years ago

How does one perform matrix-vector multiplication with a SnpArray? Currently a SnpArray lacks A_mul_B! and Ac_mul_B! typed specifically for SnpMatrix, which precludes operations like x * y for SnpArray object x and Vector{AbstractFloat} object y. The routines Acsc_mul_B! and Acs_mul_B! exist, but they call A_mul_B! and Ac_mul_B!.

klkeys commented 8 years ago

Presumably we could do A_mul_B!(output1, x.A1, b) followed by A_mul_B!(output2, x.A2, b) and output = output1 + output2 if we are not interested in standardizing anything.

Hua-Zhou commented 8 years ago

I have some code that does this. Need to merge into the package.

On Jun 17, 2016, at 3:58 PM, Kevin notifications@github.com wrote:

Presumably we could do A_mul_B!(output, x.A1, b) followed by A_mul_B!(output, x.A2, b) if we are not interested in standardizing anything.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/Hua-Zhou/SnpArrays.jl/issues/2#issuecomment-226900485, or mute the thread https://github.com/notifications/unsubscribe/AEwHsXl6zXvnXoxRY7kXru62LrhZu8Pwks5qMyZ5gaJpZM4I4ymx.

klkeys commented 6 years ago

I see that there are routines Acsc_mul_B! and Acs_mul_B! which look like A_mul_B! and At_mul_B!. I move to close this issue.

note for future refernce that Julia v0.7 does away with A_mul_B! in favor of lmul! and rmul!.

Hua-Zhou commented 6 years ago

Note current Acs_mul_B! and Acsc_mul_B! only works for real matrix, not for SnpArray yet. Reopen this issue to continue development.