Skip to contents

Changelog

dataharvester 0.1.2

This update prepares the dataharvester package for continuous integration and code coverage

What’s changed

  • removed deprecated and unused get_*() functions. Most of these functions are unused and have been superseded with download_*() functions
  • removed logtable-related functions as these are accessed directly in the Python code. There should be no reason to use these in the R code
  • removed unused config functions (config_*() and *_config()). These will be re-coded in the near future, but the removal NOW prevents the need to deprecate these functions (since they are currently unused)
  • lots of fixes to documentation, imports, suggests and formatting to meet R CMD check requirements (100%)
  • example code are not run, to prevent errors in R CMD check via GitHub Actions

What’s new

  • integration to codecov
  • added new GitHub Actions scripts to automate R CMD check and code coverage
  • new badges to README for R CMD check and code coverage

dataharvester 0.1.1

Minor bug fixes and enhancements (mostly on the dev side for continuous integration). A bigger update will come soon!

dataharvester 0.1.0

Pre-release version - dataharvester is now stable enough for pre-release (hopefully). Also fixes some bugs.

What’s changed

  • fixed error in preprocessing Earth Engine data when spectral is None - this should not happen since the spectral index calculation feature is optionals
  • fixed missing symbol in a print statement in harvest()

dataharvester 0.0.16

Quick bugfix updates.

What’s changed

  • fixed error in plotting when using harvest() if path points to a file instead of a folder

dataharvester 0.0.15

In this release we fixed a few bugs and added several vignettes for the various download functions.

What’s changed

  • fixed relative path strings not being recognised by Python modules in harvest()
  • added a demo .gif to README so that readers can preview functionality

What’s new

  • created 5 new vignettes to document layers available for downloading from the various data sources

dataharvester 0.0.14

This release focuses on data extraction from rasters and plot functions that should “simply work”. We also started to implement unit testing as the package code contains some R code.

What’s changed

  • improved message outputs when using download functions. This was important as there was some overlap between Python and R message and print outputs. This update makes the feedback (to the user) less verbose and more consistent and prepares the package for a better logging feature (which will be finalise in a future update)
  • the folder path can now be retrieved directly from an object produced by download_ee(), which is useful for plotting and other data manipulation functions which rely on extracting data from a downloaded GeoTIFF file

What’s new

  • plot() for download_*() objects, thanks to the terra package
  • extract_values() to extract values from GeoTIFF files
  • started to implement unit testing for some functions with testthat

dataharvester 0.0.13

Another quick bugfix release.

What’s changed

  • authenticating to GEE in RStudio Cloud is now more reliable when running initialise_harvester() as we have added an argument auth_mode to manually set the authentication method

dataharvester 0.0.12

This is a quick bugfix release.

What’s changed

  • fixed an error in harvest() that can occur when the infile attribute exists but contains no path, leading the function to incorrectly think that sampling data exists

dataharvester 0.0.11

What’s changed

  • implemented slightly better console feedback when downloading data (on Python side)
  • plot_rasters(): fixed plotting issue when only one raster is in folder

What’s new

  • S3 plot() methods now available with download_*() functions

dataharvester 0.0.10

This version updates the package Python modules. There are no major changes to the R code.

What’s changed

  • Messages imported from Python are a bit more organised (more to improve in a later version, but these are cosmetic changes)
  • download_dem(): arguments were not ordered like other similar functions, so we adjusted this
  • Better support for YAML config files (especially when some attributes are missing)

dataharvester 0.0.9

A fix to achieve reliable Python module imports (hopefully).

What’s changed

  • All functions that import Python modules point directly to the .py files, instead of the python folder. Hopefully fixes an internal testing error of "module not found"

What’s new

  • raster_query() samples one or more rasters based on existing sampling data to generate a sampling file of point-specifig geospatial information from all images

dataharvester 0.0.8

Big focus on making sure that dependencies work, as well as making sure that YAML config file processing workflow is more reproducible.

What’s changed

  • improved ’harvest()` for abilities to change the output download log name and plot rasters in a folder
  • initialise_harvester() can now create other environments, not just stick to r-reticulate
  • fixed errors in installing google cloud sdk on Windows machines

What’s new

  • plot_rasters() function to scan a folder recursively and plot the first band of all GeoTIFF images
  • create_yaml() function to generate template and skeleton YAML files. The function is incomplete, but it will be finalised in the next version
  • fixed encoding and missing link errors in documentation (#7)
  • fixed conda openSSL error (#8)

dataharvester 0.0.7

Another quick update to fix issues with Google Earth Engine authentication.

dataharvester 0.0.6

A quick update to authentication methods for Google Earth Engine (fingers crossed).

What’s changed:

  • 🛠: Changed how intialise_harvester() works with Google Earth Engine to be more compatible with RStudio Cloud/Binder.

dataharvester 0.0.5

This version focuses on more download options and improved validation to Python dependencies and Google Earth Engine.

What’s new:

What’s changed:

  • 🛠: Config/reticulate is no longer used to validate Python dependencies, since different package versions needed to be installed depending on the workspace environment (i.e. RStudio Desktop, RStudio Cloud, or Binder).
  • 🏎: performance and usability improvements to initialise_harvester()

dataharvester 0.0.4

This is mainly a bugfix 🪲 (and some testing) release with no new features.

dataharvester 0.0.3

What’s new:

  • ⚡: add_silo() for climate SILO downloads, not working 100% 😢
  • 📰: added examples to README

dataharvester 0.0.2

General bug fixes, added new functions that will someday replace get_*() functions.

What’s new:

  • ⚡: init_logtable() and update_logtable() to store download info
  • ⚡: download_slga() to download from SLGA
  • ⚡: download_dem() to download from DEM
  • ⚡: match_single() and match_multi() to validate args
  • ⚡: load_setttings() to read YAML file
  • 🛠: added CI to automate pkgdown deployment

What’s changed:

  • 🪲: ee_map() was not presenting map on RStudio viewer - fixed by switching to geemap.foliummap
  • 🔁: renamed ee_*() functions to *_ee() for consistency
  • 🏎: performance upgrades to initialise_harvester()
  • ⬆️️: updated Python dependencies including imports and .py files

dataharvester 0.0.1

This update will only describe package features as no one would have had the chance to report bugs.

  • initialise_harvester() which checks for Python depends and authentitcates Earth Engine
  • ee_collect() which generates an object for Earth Engine processing and downlaods
  • ee_preprocess() to perform masking, filtering, spectral index calculations and scaling
  • ee_aggregate() for basic temporal aggregation of images
  • ee_map() to visualise Earth Engine images on an interactive map
  • harvest() to perform “headless” data harvesting when paired with a config YAML
  • get_slga() for SLGA downloads
  • get_dea() for DEA downloads
  • Some initial documentation has been added to functions, but most are draft-y
  • README.md has been added
  • Initial pkgdown support added, but cannot be tested yet since repo is not public

dataharvester 0.0.0.9000

  • Added a NEWS.md file to track changes to the package.