March 2016 Quarterly Report

March 2016 Quarterly Report

Here's the spreadsheet report and regional quarterly yield map for Q1 2016.


Next entry

Previous entry

Related entries

Similar entries

Comments

  1. donal monaghan

    donal monaghan on 27/04/2016 7:30 a.m. #

    Some strange data seems to be left in, destroying the mean and sd calculations. URN 8377 in particular.

  2. Jamie Taylor

    Jamie Taylor on 28/04/2016 9:01 a.m. #

    Hi Donal,

    The anomalous data ought to be excluded from the mean calculation at the bottom of the sheet through the use of a "4 sigma" rule i.e. anything more than 4 standard deviations above the net mean is excluded from the calculation of the min, mean and max. The data is then highlighted in red to show that it has been excluded from the calculations. I leave the data in the report so that the donor can see that their results are anomalous, so there must be an issue with their data/metadata.

    In the case of URN8377 the whole row of data has been excluded and highlighted in red - are you seeing this?

    The mean of both yield and PR both look fine to me and the outlier removal seems to be working. Were you conducting your own mean calculation? If so it's important to remove extreme outliers first or use a robust statistic such as median or the robust mean.

    Jamie (Sheffield Solar Team)

Comments are closed.