Extract: Methods to extract from or assign to 'odcr' classes

In eo2cube/odcr: Accessing the Open Data Cube from R

Description

[ allows to subset an xarray object by its dimensions (see dim())

[[ allows to subset an xarray object by measurements/variables (e.g. spectral band), either named (character) or indexed (numeric).

xar_sel_time() allows to subset an xarray object by a character vector of dates/times.

[[<- allows to assign a measurement/variable (e.g. spectral band), either named (character) or indexed (numeric), to an existing xarray object.

Usage

## S3 method for class 'xarray.core.dataset.Dataset'
x[i = NULL, j = NULL, k = NULL, ...]

## S3 method for class 'xarray.core.dataarray.DataArray'
x[i = NULL, j = NULL, k = NULL, ...]

## S3 method for class 'xarray.core.dataset.Dataset'
x[[...]]

## S3 method for class 'xarray.core.dataarray.DataArray'
x[[...]]

xar_sel_time(x, query, exact_match = T)

## S3 replacement method for class 'xarray.core.dataset.Dataset'
x[[...]] <- value

## S3 replacement method for class 'xarray.core.dataset.Dataset'
x$... <- value

Arguments

x	xarray object, the dataset to be subsetted from or assigned to
i	numeric, first dimension index
j	numeric, second dimension index
k	numeric, third dimension index
...	numeric or character, index or indices by which to extract/assign (additional) elements
query	character vector, one or more time/date character vectors in the format of the time dimension of x or an abbreviated form of it (e.g. only the date component)
exact_match	logical, whether to return only exact matches (default) or to search for closest elements to each element in query using [difftime()]
value	xarray object, the dataset that should be assigned to x

Value

Subsetted xarray object

Examples

## Not run: 
library(odcr)

# connect to a database, store the Daatcube connection internally (default and recommended)
database_connect(app = "Sentinel_2")

# build a query list
lat <- 22.821
lon <- 28.518
buffer <- 0.05

query <- list(
  'time' = c('2020-01', '2020-03'),
  'x' = c(lon - buffer, lon + buffer),
  'y' = c(lat + buffer, lat - buffer),
  'output_crs' = 'epsg:6933',
  'resolution' = c(-20,20)
)

# load data and return an xarray object for a query
ds <- dc_load(query = c(product = "s2_l2a", dask_chunks = dict(), query))

# return dimensions of ds
dim(ds)

# subset/index ds by time, x, y
ds[13, 11:20, 21:30]

# subset/index ds by time
ds[13,,]

# subset/index ds by variable/measurement
ds[[2]]

# or do the same with a named variable or alias
ds[["nir"]]

# subset/index by exact time vectors
xar_sel_time(ds, c("2020-01-02", "2020-01-07"))

# or find closest times/date subsets by time vectors
xar_sel_time(ds, c("2020-01-01", "2020-01-07"), exact_match = F)

## End(Not run)

eo2cube/odcr documentation built on Dec. 20, 2021, 5:21 a.m.