search

MomX / Momocs

:dove: Morphometrics using R
http://momx.github.io/Momocs/
51 stars 19 forks source link

Alignment based on first harmonic #89

Closed TJ-Stammer closed 9 years ago

TJ-Stammer commented 9 years ago

Hello Vince,

I've tried to align based on landmarks and also have used various coo_functions to try to align my outlines to one another. Unfortunately, I keep getting large amounts of variation due to asymmetry even when I correct for it by cutting my shapes in half and using a reflection of one side to eliminate asymmetry. Whenever I change my alignment methods I get different morphospaces. What's bugging me is that my results from Shape and Momocs usually come out drastically different and I think it's due to the fact that I normalize based on the first harmonic in Shape. I was just wondering if there was a similar option in momocs or if, due to the nature of working with coo objects, it is impossible to replicate my results.

Thanks in advance! TJ

vbonhomme commented 9 years ago

when you're happy with a pipe of coo_functions, then you can pass the Out to efourier with norm=FALSE. See also the Details sections. eg.

data(bot) library(magrittr) bot %>% coo_align() %>% coo_slidedirection("E") %>% efourier(norm=FALSE) %>% PCA() %>% plot("type")

2015-07-13 19:17 GMT+02:00 TJ-Stammer notifications@github.com:

Hello Vince,

I've tried to align based on landmarks and also have used various coo_functions to try to align my outlines to one another. Unfortunately, I keep getting large amounts of variation due to asymmetry even when I correct for it by cutting my shapes in half and using a reflection of one side to eliminate asymmetry. Whenever I change my alignment methods I get different morphospaces. What's bugging me is that my results from Shape and Momocs usually come out drastically different and I think it's due to the fact that I normalize based on the first harmonic in Shape. I was just wondering if there was a similar option in momocs or if, due to the nature of working with coo objects, it is impossible to replicate my results.

Thanks in advance! TJ

— Reply to this email directly or view it on GitHub https://github.com/vbonhomme/Momocs/issues/89.

vincentbonhomme.fr http://www.vincentbonhomme.fr

TJ-Stammer commented 9 years ago

Thanks!

On Mon, Jul 13, 2015 at 6:27 PM, Vincent Bonhomme notifications@github.com wrote:

when you're happy with a pipe of coo_functions, then you can pass the Out to efourier with norm=FALSE. See also the Details sections. eg.

data(bot) library(magrittr) bot %>% coo_align() %>% coo_slidedirection("E") %>% efourier(norm=FALSE) %>% PCA() %>% plot("type")

2015-07-13 19:17 GMT+02:00 TJ-Stammer notifications@github.com:

Hello Vince,

I've tried to align based on landmarks and also have used various coo_functions to try to align my outlines to one another. Unfortunately, I keep getting large amounts of variation due to asymmetry even when I correct for it by cutting my shapes in half and using a reflection of one side to eliminate asymmetry. Whenever I change my alignment methods I get different morphospaces. What's bugging me is that my results from Shape and Momocs usually come out drastically different and I think it's due to the fact that I normalize based on the first harmonic in Shape. I was just wondering if there was a similar option in momocs or if, due to the nature of working with coo objects, it is impossible to replicate my results.

Thanks in advance! TJ

— Reply to this email directly or view it on GitHub https://github.com/vbonhomme/Momocs/issues/89.

vincentbonhomme.fr http://www.vincentbonhomme.fr

— Reply to this email directly or view it on GitHub https://github.com/vbonhomme/Momocs/issues/89#issuecomment-121079875.