Beta div failing with pathways

[asizemore]

Noticed on both live and qa.

Example (qa):

R output:

Rserve started in non-daemon mode.
[1] "starting beta diversity computation"

2023-12-05 10:41:13 Received df table with 574 samples and 435 taxa.

2023-12-05 10:41:13 Replaced NAs with 0

2023-12-05 10:41:14 Computed dissimilarity matrix.
Error in eigen(delta1) : infinite or missing values in 'x'
In addition: Warning messages:
1: no function found corresponding to methods exports from ‘veupathUtils’ for: ‘asJSON’
2: replacing previous import ‘data.table::as.data.table’ by ‘veupathUtils::as.data.table’ when loading ‘plot.data’
3: replacing previous import ‘veupathUtils::toJSON’ by ‘jsonlite::toJSON’ when loading ‘plot.data’
4: replacing previous import ‘data.table::as.data.table’ by ‘veupathUtils::as.data.table’ when loading ‘microbiomeComputations’
5: In data.table::fread("beta_div_input", select = c(OBI_0002623.WholeMetagenomeSequencingAssay_stable_id = "double",  :
  Attempt to override column 1 <<OBI_0002623.WholeMetagenomeSequencingAssay_stable_id>> of inherent type 'string' down to 'float64' ignored. Only overrides to a higher type are currently supported. If this was intended, please coerce to the lower type afterwards.
6: In data.table::fread("beta_div_input", select = c(OBI_0002623.WholeMetagenomeSequencingAssay_stable_id = "double",  :
  Attempt to override column 2 <<EUPATH_0000609.Sample_stable_id>> of inherent type 'string' down to 'float64' ignored. Only overrides to a higher type are currently supported. If this was intended, please coerce to the lower type afterwards.
7: In data.table::fread("beta_div_input", select = c(OBI_0002623.WholeMetagenomeSequencingAssay_stable_id = "double",  :
  Attempt to override column 3 <<EUPATH_0000096.Participant_stable_id>> of inherent type 'string' down to 'float64' ignored. Only overrides to a higher type are currently supported. If this was intended, please coerce to the lower type afterwards.
8: In vegan::vegdist(df[, -..allIdColumns], method = method, binary = TRUE) :
  you have empty rows: their dissimilarities may be
                 meaningless in method “bray”
9: In vegan::vegdist(df[, -..allIdColumns], method = method, binary = TRUE) :
  missing values in results

[asizemore]

note not every pathway abundance collection is failing. Just some

[d-callan]

i can confirm this is a case where we have for at least one sample a 0 abundance for all pathways. ive also confirmed that is how the data are in the files produced by the workflow. i have not looked in any depth at the workflow to know if theres something we could do about it there. in any case, i dont think we can ever guarantee we wont get data like this, so maybe our AbundanceData class needs to learn how to prune itself better.

[d-callan]

confirmed working w microbiomeComputations v3.3.5