R/data wrangling/on hold/ellison_2017b.r

In chase-lab/BioTIMEx: Aggregating long-term ecological communities data

## ellison_2017b
# correct typos in subs, method
if (FALSE) {
   dataset_id <- 'ellison_2017b'
   load(file='data/raw data/ellison_2017b/ddata')
   setDT(ddata)

   setnames(ddata, old = c('cham', 'subs', 'n'),
            new = c('site', 'block', 'value'))

   ddata[, ':='(year = format(date, '%Y'),
                month = as.numeric(format(date, '%m')),
                treat = gsub(' ', '_', treat),
                species = ifelse(!is.na(species),
                                 as.character(species),
                                 paste0(genus, ' sp.'))
   )
   ][,
     treat := gsub('chamberless', 'chamberless_cont', treat)
   ]


   warming_table <- unique(ddata[treat != 'pre-treat', c('site','treat','warming')])

   warming_table[, ':='(treatment = paste(treat, ifelse(!is.na(warming), warming, 0), sep = '_'))
   ]

   ddata <- merge(ddata, warming_table[,.(site, treatment)], by = c('site'))

   # community
   ddata[, ':='(N = sum(value),
                S = length(unique(species)),
                ENSPIE = vegan::diversity(x = value, index = 'invsimpson')
   ),
   by = .(site, block, year, treat)
   ]
   ddata[, minN := min(N), by = .(site, block)]
   #data.frame(ddata[, .(minN=min(N)), by = .(site, block, year)][order(minN)][minN<6])

   ddata[, Sn := vegan::rarefy(value, sample = minN), by = .(site, block, year)]




   ddata[, ':='(
      dataset_id = dataset_id,

      treatment_type = 'warming',
      time_since_disturbance = ifelse(treatment == 'pre-treat' | grepl(treatment, pattern = 'cham'), NA, 1),
      timepoints = paste0('T', seq_along(unique(year))[match(year, unique(year))]),
      design = paste0(ifelse(treatment == 'pre-treat', 'B', "A"),
                      ifelse(grepl(treatment, pattern = 'cham'), 'C', "I")),
      realm = 'terrestrial',
      taxon = 'invertebrates',

      comment =  'Block design with treatments being, no chamber, a chamber without warming, a chamber and warming with different warming intensities. No Winkler samples in this dataset. Repeated samplings in a single year are pooled. Counts are added and divided by effort. Effort is defined as the number of surveys per year (1 to 4).'
   )
   ]


   dir.create(paste0('data/wrangled data/', dataset_id), showWarnings = FALSE)
   fwrite(ddata, paste0('data/wrangled data/', dataset_id, "/", dataset_id, '.csv'),
          row.names=FALSE)

}