otosearch: Search new specimens against project database
In jinyung/otolith: Identification of fish otolith using digital images

Description Usage Arguments Details Value See Also

search the images/ semi-landmarks configuration(s) of new, unknown specimens against the configurations saved in the project.

1 2	otosearch(project, query, show = 5, saveresult = FALSE, showplot = TRUE, savename)

`project`	path to a project (`.rds` file) to be read, saved using `saveproj`, or a project object already read into R. if not given, interactive file selector will pop out to prompt user to select a `.rds` file (Windows only)
`query`	path(s) to otolith images/ path(s) of folder containing the otolith images/ `.tps` file containing the semi-landmark configurations/ p x k matrix or p x k x n array of semi-landmark configuration(s) to be searched. If none is given, interactive file selector will pop out to prompt user to select images to be searched (Windows only)
`show`	integer. how many search results to be shown (in order of ranking). Note that if number of specimens in database searched against is less than show, the result shown will be less than the given number
`saveresult`	logical. whether to save the result
`showplot`	logical. whether to show plot of query and their matches
`savename`	character. optional. the file name, if `saveresult=TRUE`

The search is based on the Procrustes (Riemannian) distance between the query and the database. the lower the distance (rdist) between the query and the project's configuration, the higher their ranking in the search result. Perfect match will have 0 distance (for non-sliding landmark method).

To shorten the time of searching, the query is first searched against the meanshape configurations of each species present in the project. For each species in the project, maximum distance between individuals and the species's meanshape was already calculated (by sprdist, wrapped within rGPA). First the search will determine if the query to each species's meanshape distance is within this range, if say, the query is within the range of 2 species' range, then searching will be continued with the individuals of these 2 species only. Otherwise, searching will be continued with the individuals of the top 5 closest species only.

The search has taken the semi-landmarks configuration arrangements into consideration already, hence the user doesn't need to know the side/ direction of the query.

Note: sliding semi-landmark is currently not supported due to the difficulty to achieve short searching time and correct configuration guessing at the same time. Therefore the gpa object saved in the project should be built with fixed semi-landmarks. However, this does not affect otopred, thus if sliding is preferred, otopred will still give good predictions.

The search result. if saveresult=TRUE, the result is written into a .txt file. The results include:

rdist: distance between the query and the match
label: label of the match from the database
species: species of the match
inside: whether the rdist is within the range of that species in the database
orient: type of configuration that match