Installation

Ruby + Jekyll

The website is created from markdown files rendered to html by Jekyll

• this site explains how Jekyll can be used on Windows: Run Jekyll on Windows
• download and install Ruby (64bit) from rubyinstaller.org
• add ruby path to PATH environment variable C:\Users\[Username]\Programs\Ruby21-x64
• download and install the Ruby DevKit (64bit) from the same page, search For use with Ruby 2.0 and above (x64 - 64bits only)
• install Jekyll gem install jekyll
• open a git bash terminal and navigate to Jekyll root folder
• install Jekyll dependencies bundle exec install
• serve the site locally bundle exec jekyll serve
• export contents to sws_r_share: rsync -h -v -P -t -u --recursive /cygdrive/c/Users/[Username]/path/to/_site/* //hqlprsws1.hq.un.fao.org/sws_r_share/fbsmodules

Git bash + Make

The contents of the website are created and updated using R scripts that are triggered by Makefiles

• the folder structure based on dryworkflow
• the Makefile in the root directory indicates subfolders that are checked for additional Makefiles
• the Makefile definitions for R can be found at r-rules.mk
• open a git bash terminal and navigate to dryworkflow root folder
• execute make

Modify Contents

Diagrams

The website includes a number of diagrams that facilitate understanding the project structure

• most diagrams have been created using yEd, an open source diagram editor yWorks yEd Graph Editor
• .graphml files are located at /assets/diagram/
• to include in the jekyll website, diagrams are exported from yEd in svg format
• to display in github and include in PDF documentation, png format is required
• few diagrams have been created using the online tools draw.io and Google Drawings

Visualizations

The heatmap visualization is using the R package highcharter that can only be used internally for licensing reasons

• /extra/faostat_bulk.csv specifies the datasets that are downloaded from FAOSTAT
• /downloadData/download_faostat_bulk.R downloads the data from the internet
• /data/codebook/faostat_trade_crops_livestock_norm_codebook.csv is the codebook for renaming variables
• /readMergeData/read_codebook_faostat_trade_crops_livestock_norm.R loads the codebook
• /readMergeData/read_faostat_trade_crops_livestock_norm_csv.R reads the downloaded CSV data and stores in RData format
• /readMergeData/clean_faostat_trade_crops_livestock_norm_csv.R applies the renaming specified in the codebook
• /work/summary_faostat_trade_crops_livestock_norm_csv.R subsets the data to selected commodities
• continental regions are defined in the stanDim list of objects contained in the stan package: data(stanDim)
• /fragments/fragment_faostat_trade_crops_livestock_csv.Rmd creates the html fragment for time series published on FAOSTAT

Import Contents

Once the changes have been completed in the dryworkflow routine(s), the compiled html fragments files are imported to the Jekyll site. The contents of the import-dryworkflow.bat are shown below. In this example, fragments from two dryworkflow projects are integrated: one for all FBS, the other for trade-specific outputs.

set JEKYLLFAOCYG="/cygdrive/c/Users/[User]/path/to/jekyll/site"
set DRYFAOFBSCYG="/cygdrive/c/Users/[User]/path/to/dryworkflow/root"
set SWSDATA="/cygdrive/c/SWS-Data"

rsync -h -v -P -t --recursive %DRYFAOFBSCYG%/includes/* %JEKYLLFAOCYG%/_includes/dryworkflow/
rsync -h -v -P -t --recursive %DRYFAOFBSCYG%/figures/* %JEKYLLFAOCYG%/assets/dryworkflow/
rsync -h -v -P -t --recursive %DRYFAOFBSCYG%/fragments/*.html %JEKYLLFAOCYG%/_includes/rmarkdown_fragment/

rsync -h -v -P -t --recursive %SWSDATA%/faoswsTrade/fragments/*.html %JEKYLLFAOCYG%/_includes/rmarkdown_fragment/faoswsTrade/
rsync -h -v -P -t --recursive %SWSDATA%/faoswsTrade/figures/* %JEKYLLFAOCYG%/assets/dryworkflow/faoswsTrade/