jefferislab/catnat

neuroanatomy toolbox – CATMAID extension functions

catnat's Issues

bug when there are NULL partner_skid in axonal input to be updated

install instructions still point to `alexanderbates/catnat`

Connector URLs have spaces

Problem is that

apply(df, 1, connector_URL)

turns each row of df into a character matrix.

primary.neurite.neuronlist() should pass on ...

primary.neurite.neuronlist<-function (someneuronlist, ...) 
{
    nlapply(someneuronlist, primary.neurite.neuron, OmitFailures = T, ...)
}

Documentation improvements

If anyone is going to use this package, we should try to improve the documentation a bit. I have added a pkgdown online documentation site. See e.g. https://pkgdown.r-lib.org/articles/pkgdown.html for config details.

You will see there are currently lots of duplicate doc errors (some functions are defined 3x) and also a bunch if incorrectly documented args. it would be good to fix.

primary.neurite.neuronlist is not called for neuron lists

the generic primary.neurite function should look like this

primary.neurite<-function(x, ...) UseMethod("primary.neurite")

this picks the appropriate specialised method according to the type of x

You can also have a catch all

primary.neurite.default

function which may be what you intended for:

primary.neurite<-function (someneuron, k = 100) 
{
    som = soma.neuron(someneuron)
    if (is.na(som[1])) {
        if (length(someneuron[[1]]$tags$soma[[1]]) > 0) {
            som = matrix(xyzmatrix(someneuron)[someneuron[[1]]$d$PointNo %in% 
                someneuron[[1]]$tags$soma, ], ncol = 3)
        }
        else {
            som = matrix(xyzmatrix(someneuron)[someneuron[[1]]$StartPoint, 
                ], ncol = 3)
        }
    }
    p = nat::xyzmatrix(someneuron)
    n = nabor::knn(p, som, ifelse(nrow(p) > k, k, nrow(p)))
    m = p[c(n$nn.idx), ]
}

get_connected_skeletons behaves unexpectedly

Seba encountered this: get_connected_skeletons appears to return incomplete list of partners.

Minimal example:

> av1a1Neurons = read.neurons.catmaid(c(2659704, 2684792))
> cn = get_connected_skeletons(av1a1Neurons, min_synapses = 10)
> names(cn)
[1] "2101395"  "4947032"  "6180224"  "6189928"  "6199308"  "6199551"  "11545803"

cn should contain order 50 neurons given a 10 synapse threshold.

Bug in resample.catmaidneuron for neurons without connectors

> asc_92902213 = read.neurons.fafbseg(92902213)
> asc_92902213.resampled1 = resample(asc_92902213, stepsize = 10)
Error in array(x, c(length(x), 1L), if (!is.null(names(x))) list(names(x),  : 
  'data' must be of a vector type, was 'NULL'
> asc_92902213.resampled2  = nlapply(asc_92902213, resample, OmitFailures = FALSE, stepsize = 10)
> plot3d(asc_92902213.resampled2)
Warning message:
In xyz.coords(x, y, z, xlab = xlabel, ylab = ylabel, zlab = zlabel,  :
  NAs introduced by coercion

nat.nblast must be a dependency

> examples = primary.neurite.tracts = nlapply(pnts,function(z) average.tracts(subset(most.lhns.pnts,pnt==z),mode=1))
  |=================================================================| 100%
  |=================================================================| 100%
 Show Traceback
 
 Rerun with Debug
 Error in get(smat) : object 'smat.fcwb' not found 

This is a subtle namespace that @Ajdm2 and greg never solved.

googlesheets dependency is archived

This won't be a very common problem at this point but you might want to consider switching the google sheets related functions to use googlesheets4 instead of googlesheets, which is archived: https://github.com/jennybc/googlesheets

Right now the dependency on googlesheets makes installing catnat a bit inconvenient.
i.e. one has to get googlesheets from the CRAN archive:
install.packages("https://cran.r-project.org/src/contrib/Archive/googlesheets/googlesheets_0.3.0.tar.gz", repos = NULL, type = "source")

nat.nblast not nat.blast in average tracts

> examples = primary.neurite.tracts = nlapply(pnts,function(z) average.tracts(subset(most.lhns.pnts,pnt==z),mode=1))
  |=================================================================| 100%
  |=================================================================| 100%
 Show Traceback
 
 Rerun with Debug
 Error in loadNamespace(name) : there is no package called ‘nat.blast’ 

Improve docs for catnat:::primary.neurite.neuron

In particular point out that it just finds the first segment from the root – even if it's terminated by a tiny one segment branch. Could cross-reference elmr::simplify_neuron and give an eaxample of using that.

skeleton_connectiviy_matrix only accepts neuronlist, connectivity_matrix doesn't return matrix

skeleton_connectivity_matrix
I assume it was intended to work like this, but it would be nice to have the option to use skids only without the need to fetch neurons, alternatively add

if (!(is.neuronlist(pre))) {
        stop("Neuronlist expected")
    }

connectivity_matrix:
always returns NULL, m is missing from the end of the function