Wednesday, 26 March 2014

Statistics

Completing my 245th batch of bread today I was moved to wonder about how steady the bake rate was, and so consulted the elaborate record, in the form of an Excel spreadsheet.

First step is to simply graph the dates in column 1 which yields a very straight looking line. One might deduce that the bake rate was steady.

But probing, I find 95 bakes in the first year (January 2011 to January 2012), 55 in the second and 76 in the third, with some over in what will be the fourth. Probing further, I find a number of small bakes in the first year and what looks like a gradual increase in the size of loaf - not that size is readily available as I do not record the weight of the finished bread and I record the weight of flour (a reasonable proxy) in text, in pounds and ounces, which Excel does not understand. It would take a few lines of VB code, a small program, to convert from that to ounces simple.

Then I need to do something about date, maybe numbering the months since January 2011 and assigning each bake to one of those sorts of months. A few more lines of VB code to get from date to month number. Or would day number be better?

And then I might be in business.

Which all goes to show, if one wants to get the statistics one wants, the spreadsheet needs to be designed with them in mind, rather than just knocking the thing up on the screen. I ought to have known better!

Spreadsheet to be found at https://dl.dropboxusercontent.com/u/8152054/Bread-20110120.xls.

No comments:

Post a Comment