In weecology/MATSS: Macroecological Analyses of Time Series Structure

IS_LOCAL <- identical(Sys.getenv("LOCAL"), "true")

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>", 
  purl = IS_LOCAL
)

library(dplyr)
library(tidyr)
library(purrr)
library(MATSS)

Overview

The basic info we want for each dataset is:

• taxon data (species-level if appropriate)
• sampling start and end times
• sampling frequency
• geographical location
• life history characteristics of the organisms
• temperature at sampling location

For the latter two, it's probably easier to rely on external packages to retrieve the info, so the summaries will first focus on the remaining entries.

BioTime summaries

Define a function for how to generate the summary for BioTime datasets. Note that we use a tibble structure, to enable easy assembly across datasets.

summarize_biotime_data <- function(dataset_id = 10, ...)
{
    dat <- get_biotime_data(dataset_id = dataset_id, 
                            ...)

    summ_info <- tibble(dataset_id = dataset_id)

    # determine species info
    summ_info$species_table <- I(list(dat$metadata$species_table))

    # determine start and end time
    if (dat$metadata$is_annual_sampling)
    {
        var <- "year"
    } else {
        var <- "date"
    }
    summ_info$start_time <- as.character(min(dat$covariates[[var]], na.rm = TRUE))
    summ_info$end_time <- as.character(max(dat$covariates[[var]], na.rm = TRUE))

    # determine sampling frequency
    if (dat$metadata$is_annual_sampling) # missing month data in raw values
    {
        summ_info$sampling_frequency <- "annual (or less)"
    } else if (length(unique(dat$covariates$year)) >= NROW(dat$covariates)) {
        summ_info$sampling_frquency <- "annual (or less)"
    } else {
        summ_info$sampling_frequency <- "subannual"
    }

    # determine spatial extent
    summ_info$number_locations <- dat$metadata$number_lat_long
    summ_info$lat <- dat$metadata$cent_lat
    summ_info$long <- dat$metadata$cent_long
    summ_info$lat_min <- min(dat$covariates$latitude, na.rm = TRUE)
    summ_info$lat_max <- max(dat$covariates$latitude, na.rm = TRUE)
    summ_info$long_min <- min(dat$covariates$longitude, na.rm = TRUE)
    summ_info$long_max <- max(dat$covariates$longitude, na.rm = TRUE)
    return(summ_info)
}

Results

Now repeat for all the datasets within BioTime

biotime_dataset_ids <- get_biotime_dataset_ids()

biotime_summaries <- map_dfr(biotime_dataset_ids, summarize_biotime_data)

Extract out just the dataset-level info, and save it for the vignette to be built without needing to download and process the BioTime data first.

dataset_info <- biotime_summaries %>% 
    select(-species_table)

saveRDS(dataset_info, here::here("inst/biotime_dataset_info.RDS"))

dataset_info_file <- system.file("biotime_dataset_info.RDS", 
                                 package = "MATSS", mustWork = TRUE)
dataset_info <- readRDS(dataset_info_file)
knitr::kable(dataset_info, digits = 2)

Extract out just the taxa:

species_info <- biotime_summaries %>%
    unnest() %>%
    select(dataset_id, genus, species, genus_species)

# knitr::kable(species_info)

Search for duplicates

weecology/MATSS documentation built on May 15, 2020, 7:03 p.m.