Updates needed to workflows

[jdechalendar]

• Update Makefile to make bulk dataset easier to reproduce -- see Pipeline updates (and docs) #47
• Update workflow to save deltas here -- done in Updates to workflows #44
• Update docs on different types of datasets #40
• Merge six month datasets -- done in Various updates to plotting utils and dataset convenience tools #42
• Extra timesteps: are these added by the rolling window cleaner? Possibly related to Bulk Processed Files - dealing with duplicate periods across years #30
• Store gridemissions version in name of dataset that is uploaded to aws (processed.v0.1.7.tar.gz). Check version on load. Also store date it was created? -- Done in 1c8a883
• Add a parquet file with the full dataset for users who just want to deal with one file
  • Consider merging the six month files into one dataset at the beginning of the workflow and running that

[jdechalendar]

Extra timesteps: this does not seem to show up for the live dataset. Need to look again at what is happening for the bulk dataset