start_annotation_simple: Start simple annotation

In rstoat: R Interface to the 'STOAT' Platform

Description

Submit a dataframe for on-the-fly annotation. Does not require login - for use for small numbers of records and pilot jobs.

You can annotate using layers that are on earthengine! Layer parameters: Used by all:

• spatial: The spatial buffer in meters.

• temporal: The temporal buffer in days.

For STOAT layers:

• product: The product e.g "srtm", or "landsat8".

• variable; The vairable, e.g "elevation", or "evi".

For non STOAT layers, ie those in Google Earth Engine:

• id: The id of the image in Google Earth Engine, mapped to "product" in the output.

• static: Whether to load the imagery as an ImageCollection or as an Image.

• bands: A list, wiht one element, which is used to specify which band of the imagery to use for the annotation, mapped to "variable" in the output.

• reducers: A list containing one or more of the following:

  • mean

  • lcv_count

  • mode

  • median

  • stdev

  • min

  • max

  • stdev

  If only one is provided the output will be assigned to "value", else the output will be named the same as the reducer.

Usage

start_annotation_simple(
  events,
  layers,
  coords = c("lng", "lat"),
  date = "date"
)

Arguments

events	A data.frame for on the fly annotation
layers	A list of parameters or vector of codes, of the layers, see the examples below.
coords	A vector of length 2 containing column names for record longitudes, and latitudes.
date	Column name for record dates, dates must take the format YYYY-MM-DD

Value

Input data.frame with values from the annotation appended, in addition to unique identifier field event_id.

• event_id: A unique identifier for each occurrence

• product: Product used for annotation

• variable: Variable used for annotation

• s_buff: Spatial buffer in meters applied to occurrence

• t_buff: Temporal buffer in days applied to occurrence

• value: Annotated value of occurrence from requested layer (mean within buffer), if there is only one reducer (default), then this value will be here.

• stdev: Standard deviation of values within buffer

• valid_pixel_count: Number of pixels within buffered area'

Examples

## Not run: 
events <- data.frame(
   event_id = as.character(1:2),
   lng = c(-4, 24),
   lat = c(10, 10),
   date = '2015-01-01'
)

# simple layer string format: PRODUCT-VARIABLE-S_BUFF-T_BUFF
layers <- 'landsat8-evi-100-16'
start_annotation_simple(events, layers)

start_annotation_simple(events, layers)

# For lcv_count (Count of landcover value), 'value' returned is a string of
# landcover counts within the AOI.
# The output format is:
#  <LANDCOVER_CLASS>:<COUNT_OF_PIXELS_WITH_THAT_CLASS>
#  classes are seperated by commas.

start_annotation_simple(events, list(
  list(
    id="COPERNICUS/Landcover/100m/Proba-V-C3/Global",
    s_buff=1000,
    reducers=list("lcv_count", "mode"),
    static=FALSE,
    t_buff=365,
    bands=list("discrete_classification")
  )
))

# Annotating with two worldclim layers:
#   bio01 is annual mean temperature
#   bio12 is annual precipitation

start_annotation_simple(events, list(
 list(
   "id"= "WORLDCLIM/V1/BIO",
   "s_buff"=1000,
   "reducers"=list("mean"),
   "static"= TRUE,
   "t_buff"= 1,
   "bands"=list("bio01")
 ),
 list(
   "id"= "WORLDCLIM/V1/BIO",
   "s_buff"=1000,
   "reducers"=list("mean"),
   "static"= TRUE,
   "t_buff"= 1,
   "bands"=list("bio12")
 )
))

## End(Not run)

rstoat documentation built on Nov. 10, 2021, 9:07 a.m.