sequence paired identity large than 100%

[Chenglin20170390]

Hi , I test with follow data, but the sequence identity of matrix large than 1 . I don't know how to explan the result...
./famsa -dist_export -pid -square_matrix test.fa pid.csv
`>P
MMMMMRRRRR

T
MMMRRRRRRR
E
RRRRRRRRRR
F
RRRRRRRRRR`

output

,P,T,E,F
P,10000.000000,2.000000,0.500000,0.500000
T,2.000000,10000.000000,1.166667,1.166667
E,0.500000,1.166667,10000.000000,10000.000000
F,0.500000,1.166667,10000.000000,10000.000000

[agudys]

@Chenglin20170390
Thank you for raporting the issue. Indeed, in -pid mode the matrix contains inverse of the dissimilarities which may result in such strange values. We will fix it in the next release to make sure that identities are from [0,1] interval.

[agudys]

@Chenglin20170390
In the latest (2.2.1) version pairwise identity is calculated as the number of matching residues divided by the length of the shorter sequence. Please let me know if everything works as expected.

Adam