Specify minimum R version in `README`

michaelmhoffman commented 3 years ago

I just tried to install spiky and after 15 min 6 s of installing dependencies get

The downloaded source packages are in
        ‘/tmp/RtmpVyzEvE/downloaded_packages’
✔  checking for file ‘/tmp/RtmpVyzEvE/remotes2316accfa18/trichelab-spiky-e0cef94/DESCRIPTION’ ...
─  preparing ‘spiky’:
✔  checking DESCRIPTION meta-information ...
─  checking for LF line-endings in source and make files and shell scripts
─  checking for empty or unneeded directories
─  looking to see if a ‘data/datalist’ file should be added
─  building ‘spiky_0.99.12.tar.gz’ (2.6s)

Installing package into ‘/usr/local/lib/R/site-library’
(as ‘lib’ is unspecified)
ERROR: this R is version 3.6.3, package 'spiky' requires R >= 4.0.0
Error: Failed to install 'spiky' from GitHub:
  (converted from warning) installation of package ‘/tmp/RtmpVyzEvE/file2315d40c102/spiky_0.99.12.tar.gz’ had non-zero exit status
Execution halted

This is frustrating, could you please mention the minimum R version in the README?

michaelmhoffman commented 3 years ago

Is the minimum of 4.0.0 really necessary for this package? I can't even get this through Conda which only provides 3.6.1 (or 3.5.0 for Linux-x32). I can see that the latest Bioconductor requires 4.0.3.

ttriche commented 3 years ago

I can check. I haven't attempted to run with anything less than 3.6.3 (and personally I despise conda which always seems to wreck my installations of... everything unless individual conda envs are parceled out for each little task). Supporting older versions of R may or may not be worth the trouble given that Bioconductor distribution implies >= 4.0.0. But I'll give it a shot on my old beater laptop and see what happens.

ttriche commented 3 years ago

> R.version$version.string
[1] "R version 3.6.3 (2020-02-29)"
> model_glm_pmol(scan_spike_counts(sb, spike=spike))

Call:  glm(formula = conc ~ read_count + fraglen + GC + CpG_3, data = x)

Coefficients:
(Intercept)   read_count      fraglen           GC        CpG_3  
  9.000e-04    5.152e-24    1.353e-20   -6.435e-20   -4.052e-19  

Degrees of Freedom: 25 Total (i.e. Null);  21 Residual
Null Deviance:      3.056e-37 
Residual Deviance: 5.689e-36    AIC: -2109
> model_bam_standards(scc)
AICc -3643.25 logPost 1780.803 logLik 1827.763 edf 6.0000 eps 0.1952 iteration  AICc -3666.77 logPost 1792.565 logLik 1839.525 edf 6.0000 eps 0.0118 iteration  AICc -3667.70 logPost 1793.031 logLik 1839.991 edf 6.0000 eps 0.0025 iteration  AICc -3667.70 logPost 1793.032 logLik 1839.992 edf 6.0000 eps 0.0001 iteration  AICc -3667.70 logPost 1793.032 logLik 1839.992 edf 6.0000 eps 0.0001 iteration   4
elapsed time:  0.01sec
|********************| 100%  0.00sec  1.57sec
Family: gaussian 
Link function: mu = identity, sigma = log
*---
Formula mu:
---
conc ~ read_count + fraglen + GCfrac + CpGs_3

Formula sigma:
---
sigma ~ 1

I guess it runs with R-3.6.3, or at least tolerates it. You're going to have a bad time with CRAMs on 3.6.3, though.

trichelab / spiky

Specify minimum R version in `README` #16