- quotes for coverage between $11,500 and $19,500
- split into low and high coverage segments
- linear regression
- plots quotes and best fit for both segments
Premium vs Coverage
- historical weather data: 1943-present, minimum/maximum daily temperatures (source : BOM)
- cleaned data, merged data1 and data2, calculated daily difference value (max - min)
- created rolling means, std for each series
- plotted results
- increase in both max and min over time period, however daily difference difference is steady
- daily difference std is as expected at sqrt(max_std^2 + min_std^2) / sqrt(2)
- normalise std series by mean series, what is the correlation?
- show Bollinger shading (mean plus/minus 2*std)
- break max/min into seasons, is there a difference in trend at season level?
- changes to hottest/coldest tme of year? earlier/later?
Daily min (green), max (black) and differnece (red) rolling (window = 4 years) mean (first) and std (second)
