Visualizing Time Series Data
- Please see 'TimeSeries Markdown.md' document above for full analysis
- Also available on RPubs at https://rpubs.com/brosnahj/653877
We have been asked by a law firm to conduct an in-depth analysis of power consumption data for a client's residential home. The law firm's client claims to have not been occupying a specific residence at the time of an undisclosed event during Summer of 2008. Energy use records will be used to provide evidence on whether or not residence was occupied from July-September 2008.
Our objectives are to conduct an in-depth analysis of energy records from 2007 to 2010 for client residence and to answer the law firm's question, 'Was client residence occupied during the Summer of 2008?' We will achieve this by visualizing energy use patterns on high level (up to 3-years) to determine overall energy use patterns as well as a microscopic level during Summer of 2008
- Time Series data and preprocessing
- Tidyverse data wrangling
- ggplot2 time series visualizations
- Deep analytics
- Business objectives achieved
- Seasonal use patterns reveal peak energy use in winter and low in summer months, with water heater & AC consistently using more across time
- There was a sharp, extended decline for all sub-meters seen from Aug 5th - 31st, 2008
- This decline is atypical of all other time periods
- No energy was used from kitchen submeter from Aug 6th - 31st, 2008
- Lower energy use than typically seen in summer from both Laundry and Water Heater/AC submeters
- Recommendation based on evidence from data is that residence was not occupied from August 6th - August 31st, 2008, but was occupied all other days within July-September 2008 time frame.