1 Dec Webcast - Methods for Quickly Shaping Data for Cluster and Time Series Analyses
Created:
| Last Modified:
In this webcast we will start with a .CSV file downloaded from the Department of Energy - Energy Information Administration (DOE - EIA), import it into JMP with the Excel Import Wizard, then stack and split columns, and then stack columns again as we "shape" the data for various analyses. In particular, for cluster analysis we will want the data "wide" and for time series analysis we will want the data "tall."
Data tables to be used are attached at the bottom of this page and include the .CSV file downloaded using the DOE EIA website's "Electricity Data Browser." Average Retail Price of Electricity data are provided from January 2001 through August 2013 for the 50 states + DC, grouped by regions of the country, and provided for three industry sectors plus their total. All intermediate files along the way have been saved so that attendees can download them ahead of time and follow along during the webcast. The emphasis in this webcast will be the mechanics of stacking and splitting the data to facilitate the cluster and time series analysis shown. The ease of analyses tend to be the "wow" of JMP, but the mechanics of shaping the data will sometimes elicit comments like "What you just did in minutes in JMP took me days in my spreadsheet!"
Below are several graphs relating to the clustering of the 50 States + DC by Average Retail Electricity Price over 12+ years.
Below are two graphs showing time series forecast for Average Retail Electricity Price in California.
NOTE: A single 0.8 Mb file named "01 Dec Webcast DOE-EIA Files.zip" contains the 16 separate files that will be used in this webcast. Since not all users are allowed to download *.zip files, we are posting the files separately as well as zipped. There is a limit of ten attachments to this webpage and so 9 of the 16 files are attached below. The remaining 7 files are attached to the page at this link.
