NEWS.md

In febr: Data Repository of the Brazilian Soil

News

febr 1.9.9

• Bug fix. There were R code comments using non-ASCII characters. This was causing a warning in Debian Linux, R-devel, clang, ISO-8859-15 locale. Those comments were removed in this patch.

febr 1.9.8

• Improvement. Package description is clearer and follows CRAN policies.

febr 1.9.7

• Bug fix. Data downloaded using identification(), layer(), metadata(), and observation() were supposed to be named after the respective data id used in the repository. However, the object passed to names() was incorrect. This is corrected in the present patch.

febr 1.9.6

• Bug fix. Data downloaded using readFEBR() were supposed to be named after the respective data id used in the repository. However, the object passed to names() was incorrect. This is corrected in the present patch.

febr 1.9.5

• Improvement. The data dictionary of the FEBR repository is updated periodically and some fields are inactivated. The function dictionary() has a new argument to enable the user to select whether active (active = TRUE), inactive (active = FALSE) or any (active = NULL, default) field should be returned.

febr 1.9.4

• Improvement. The data dictionary of the FEBR repository is updated periodically. Some data field names are hard-coded in the febr package. This minor version accounts for some of these changes, specifically:
• coord_datum -> coord_datum_epsg

febr 1.9.3

• Improvement. The data dictionary of the FEBR repository is updated periodically. Data field names need to be updated upon data download. Until recently, only the names of mandatory data fields were updated. Now, the name of any data field is updated.

febr 1.9.2

• Improvement. The data dictionary of the FEBR repository is updated periodically. Recently, the term evento (event) was adopted as a replacement for observacao (observation). Some data field names are hard-coded in the febr package. This minor version accounts for some of these changes, specifically:
• observacao_id -> evento_id_febr
• observacao_data -> evento_data

febr 1.9.1

• Improvement. Minor documentation changes, making it more accurate. The repository is now called Data Repository of the Brazilian Soil.

febr 1.9.0

• Improvement 1. New function readVocabulary(). This function reads the table containing the controlled vocabulary used in the FEBR Soil Data Repository. The controlled vocabulary is used in FEBR to standardize the values of categorical soil variables.
• Improvement 2. Expands the documentation of function readIndex().
• Improvement 3. Adds new author (Taciara Zborowski Horst) and contributor (Glauber José Vaz). Modifies the package description, making it more accurate: the repository is now called 'FEBR Soil Data Repository'.

febr 1.8.1

• Improvement 1. Documentation on how to install the development version from GitHub was expanded. This was necessary because some users of Windows machines were having trouble dealing the installation of Rtools.
• Improvement 2. The GitHub repository was transferred to Laboratório de Pedometria. The URL, however, had not been updated in goto(), thus taking the user to the old GitHub repository. The repository URL was updated accordingly.
• Bug fix. Both layer() and observation() had examples for how to read the data files from a local directory. This is not an option for most users and was creating confusion. Examples were removed from both functions.

febr 1.8.0

• Improvement 1. Mandatory fields for data tables are not hard coded in the package anymore. Instead, now they are read directly from the online repository dictionary. This enables any update in the dictionary to have an imediate effect for users. Before this minor release, changes in the dictionary required the package to be updated, generally resulting in some unwanted delay.
• In the dictionary, mandatory fields are now indicated using a new field called campo_vital.
• Improvement 2. The online repository dictionary is reviewed and expanded periodically. Some changes include the modification of field names. These modifications can affect the functioning of functions such as observation() or layer(). To avoid these negative impacts, now these functions contain code to update field names before executing any data processing step e.g. standardization and harmonization.
• In the dictionary, old field IDs are now recorded using a new field called campo_oldid.
• Improvement 3. The package documentation has been updated accordingly.

febr 1.7.5

• Bug fix. Corrects data table stacking when embargoed datasets are present. The previous version ignored that embargoed datasets return a string when called using observation() or layer(). The presence of strings broke the data table stacking procedure.

febr 1.7.4

• Bug fix. Corrects the evaluation of stack = TRUE when all datasets data.set = "all" are read using observation() and layer(). The previous version ignored that data.set = "all" returns a vector with hundreds of datasets, and thus stack was automatically set to stack = FALSE.

febr 1.7.3

• Bug fix. Corrects the construction of file paths when all datasets data.set = "all" are read using readFEBR(). The previous version constructed the file paths using the string passed to data.set in the function call i.e. "all" instead of the vector of dataset IDs.

febr 1.7.2

• Shortens examples to avoid CRAN notifications.

febr 1.7.1

• Code and documentation clean up.
• Packages dplyr and glue are not dependencies anymore.

febr 1.7.0

• Code and documentation clean up.
• Defuncts obsolete function: dataset, febr, febr2spdf, febr2xlsx, standard, header.

febr 1.6.4

• Start using the data.table package for faster data processing.
• Code and documentation cleanup.
• New utility function readIndex() for reading the TXT file listing all data sets published in the FEBR.

febr 1.6.3

• morphology: deals with depth data between parenthesis when retrieving soil color data.

febr 1.6.2

• morphology: deal with varying number of soil color records (1 or 2)
• Documentation updates.

febr 1.6.1

• morphology: retrieve soil color data when accents are missing ("umido" or "umida")

febr 1.6.0

• febr2sse: updates nomenclature (SmartSolos Expert and sse); process soil consistency data
• morphology: retrieve soil consistency data

febr 1.5.7

• Patch: guarantee that rmarkdown is up to date to build vignettes.

febr 1.5.6

• Cleans code and documentation.

febr 1.5.5

• Sets a new file to store the ID of the datasets downloaded using the core function readFEBR().

febr 1.5.4

• New function dictionary()
• Renames standard()
• Argument table has two new valid values: metadado and versionamento.

febr 1.5.3

• Code formatting and clean up.
• Improvements on how unitless variables are dealt with.

febr 1.5.2

• Code formatting and clean up.
• Improves documentation.
• Adds NA option (#N/A).
• Corrects bug in unit conversion.

febr 1.5.1

• Improves documentation.

febr 1.5.0

• New function:
• taxonomy(). Extract and process soil taxonomic data from textual soil classification description. A vignette accompanies the new function.
• Bug fix:
• observation(). Corrects an issue in argument checking.
• Fixes invalid URLs.

febr 1.4.0

• New function morphology to extract and process soil morphological properties from field soil morphology descriptions.
• Adds data conversion routines in febr2smartsolos for the following soil variables: taxonomic classification, soil matrix color (wet and dry), soil structure.

febr 1.3.2

• Improves translation of variable names when converting soil profile data between FEBR and SMARTSolos.

febr 1.3.1

• Deals with suggested packages.

febr 1.3.0

• New functions febr2smartsolos() and smartsolos2febr() for soil profile data conversion between FEBR and SMARTSolos.

febr 1.2.4

• Improves documentation for new release.

febr 1.2.3

• Improves documentation for new release.

febr 1.2.2

• readFEBR(): improves output object.

febr 1.2.1

• Improves package documentation.
• readFEBR(): new general purpose function to download raw data.

febr 1.1.2

• Bug fix: correct file format i.e. TXT instead of CSV.

febr 1.1.1

• Improvements:
• Includes changes to download data from new webserver
• New function argument febr.repo allows user to set where the data should be read from: the remote web server or a local directory
• Other changes:
• header() is now deprecated due to the new file structure used in FEBR. Users can now rely on the already existing metadata() to get the same results
• febr2spdf() is now defunct.

febr 1.1.0

• Improvements:
• Packages googlesheets, googlesheets4 and readr are not dependencies anymore;
• A new function febr2sf is available as a replacement for febr2sp -- sp is not dependencies anymore;
• febr2xlsx is now deprecated -- xlsx is not dependencies anymore;
• Bug fix:
• layer: standard variables are only set to type character when stack = TRUE.

febr 1.0.3.9004

• Getting rid of googlesheets4.

febr 1.0.3.9003

• Using Sys.sleep(time = 10) to pass checks in functions examples. This is necessary because the new Google Sheets API has a limit of 500 requests per 100 seconds per project, and 100 requests per 100 seconds per user (https://developers.google.com/sheets/api/limits). This issue has also been found elsewhere, e.g. https://stackoverflow.com/questions/53765222/python-google-sheets-api-limit-429-error-with-loop. We soon will remove the dependency upon googlesheets4 and thus solve this issue.
• Preparing for removal of googlesheets4 from the list of imports.

febr 1.0.3.9002

• Improvements:
• New function: febr2sf as a replacement for febr2sp;
• Replaces googlesheets with .readGoogleSheets;
• febr2xlsx is now deprecated and febr does not depend upon xlsx anymore;
• Improves documentation;
• Bug fixes:
• layer: standard variables are only set to type character when stack = TRUE;

febr 1.0.3.9001

• Improvements:
• Replaces sp with sf;
• Exports new data reading function: .readGoogleSheets;
• Improves documentation;
• Bug fixes:
• metadata: get correct file id.

febr 1.0.3.9000

• Replacing googlesheets with googlesheets4.

febr 1.0.3

• New version (patch) including internal changes in download functions and documentation improvements.

febr 1.0.2.9001

• Preparing for future modifications in data storage in febr. Now the sheet name (dataset, observacao, camada or metadado) is explicitly specified withing download functions. The user should not have to worry about this change.

febr 1.0.2.9000

• Bug fix: setting argument missing = list(coord = 'drop') in function observation() now correctly drops observations missing spatial coordinates. Thanks to Edberto Moura Lima, from the Federal University of Santa Maria, who identified the bug. Equivalent modifications were implemented in function layer().

febr 1.0.2

• New version (patch) -- including several bug fixes.

febr 1.0.1.9003

• Bug fixes:
  • knitr was moved from Imports to Suggests.
  • observatio(): checking if the number of observations with date is greater than the number of observations without date.
  • unit(): checking if the value passed to an argument is valid.

febr 1.0.1.9002

• #VALUE! now is considered as NA. This can be found when a calculation is performed in one of the Google Sheets.

febr 1.0.1.9001

• tr now is considered as NA. This is frequently found in the camada table.

febr 1.0.1.9000

• Bug fixes:
  • Many layers from datasets that have not gone through revision yet are lacking the identification code camada_id, which is automatically set to NA_character_. In these cases, when merging repeated layer data using layer(standardization = list(repetition = "combine")), the entire observation would be merged. This was because for all layers camada_id = NA_character_, that is, camada_id was duplicated. Now duplicates are checked using duplicated(incomparables = NA) to avoid incorrect merges.
  • The lowermost limit of some layers is recorded as being a wavy or irregular layer trasitions. This is done using /. The depth data of these layes may include a plus sign indicating that the same material can be found deeper in the soil. This poses a difficulty when processing the plus sign with eval() and parse() because / is interpreted as a division sign. Now the presence of / is checked before deciding how to deal with the plus sign.

febr 1.0.1

• New version (patch) -- including several bug fixes.

febr 1.0.0.9009

• Adjustments for new release.

febr 1.0.0.9008

• Bug fixes:
  • Corrects the arrangement of the columns of the table containing measurement units and fields names.
  • Corrects the stacking of measurement units and fields names.
  • Takes into account that both '<' and '< ' are used to indicate 'less than'.
  • - is used as a surrogate measurement unit for variables that are unitless.

febr 1.0.0.9007

• Bug fix: proceeds with standardization and harmonization only when the dataset has data on the selected variables. This should produce a minor speed improvement.

febr 1.0.0.9006

• Bug fix: - is used as a surrogate measurement unit for variables that are unitless. This is necessary to perform data standardization.

febr 1.0.0.9005

• Bug fix: exports correct data type when combining categorical data values when there are repeated measurements.

febr 1.0.0.9004

• Bug fix: stops execution of layer and observation when downloading a single dataset with stack = TRUE.

febr 1.0.0.9003

• Accommodates changes in febr:
• Field camada_numero now is camada_id.
• Field amostra_codigo now is amostra_id.

febr 1.0.0.9002

• Bug fix: during the standardization of measurement units, now understands that measurements units are stored in the second line of tables camada and observacao.

febr 1.0.0.9001

• Bug fixes.
• Updates authorship information.
• Adds METACRAN package download badge (https://www.r-pkg.org/services).
• Adds custom package devel version badge (based on https://github.com/GuangchuangYu/badger).

febr 1.0.0.9000

• New string used to identify comments in data tables camada and observacao. Now using #metadado> -- as a replacement of #unidade. This allows setting multiple lines with comments.
• Functions layer and observation: the measurement unit, field_unit, and true variable name, field_name, are exported as attributes of the output data.frame object.

febr 1.0.0

• Submission to CRAN.

febr documentation built on March 18, 2022, 5:57 p.m.