I think I might have found a bug in rhdf5. Namely, calling the following in a session where library(rhdf5) has *not* been run:
`rhdf5::h5write(donors, file = output_file, name = "sampleID")`
(where "donors" is a character vector) gives the following error:
```
Error in UseMethod("h5write") :
  no applicable method for 'h5write' applied to an object of class "character"
```

However calling the same thing after calling library(rhdf5) works fine. This was surprising to me. I could be missing something, but looks like a bug?

Session Info:
```
> sessionInfo()
R version 3.3.2 (2016-10-31)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Red Hat Enterprise Linux Server 7.2 (Maipo)

locale:
 [1] LC_CTYPE=en_GB.UTF-8       LC_NUMERIC=C
 [3] LC_TIME=en_GB.UTF-8        LC_COLLATE=en_GB.UTF-8
 [5] LC_MONETARY=en_GB.UTF-8    LC_MESSAGES=en_GB.UTF-8
 [7] LC_PAPER=en_GB.UTF-8       LC_NAME=C
 [9] LC_ADDRESS=C               LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_GB.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

loaded via a namespace (and not attached):
[1] zlibbioc_1.20.0 readr_1.0.0     assertthat_0.1  R6_2.2.0
[5] tools_3.3.2     pbdZMQ_0.2-5    rhdf5_2.18.0    tibble_1.2
[9] Rcpp_0.12.8
```

Best

Davis

I think this is because h5write isn't registered as an S3method.  When you haven't attached rhdf5 using library(), R looks for the appropriate method in the environment, and then in the database of registered methods, but it won't find it in either and fails.  Once you've run library() the h5write.default method can be found and it works.

I think this can be fixed by changing export(h5write.default) to S3method(h5write, default) in the packages NAMESPACE file.