Create a full end-to-end example for one dataset

[lawinslow]

A great starting model would be metab.bookkeep.

The code would

1. Load data
2. supply data to a model
3. retrieve and plot the output metabolism parameters.

To see how to load data out of the "inst/extdata" folder, see the timeseries function, like ts.wedderburn.number, in rLakeAnalyzer.
http://cran.r-project.org/web/packages/rLakeAnalyzer/rLakeAnalyzer.pdf

A great place to put it would be here:
https://github.com/GLEON/LakeMetabolizer/blob/master/inst/manuscript.script.R

[rBatt]

I've done some of this (see ryanData.R) – I suggest we load .RData files with the package for the demos. The .RData files would, e.g., have objects containing a matrix of 1 day of data from each lake, and maybe a series of days of data from 1 of the lakes.

[lawinslow]

I'd like to load from text files as opposed to RData files. It's slower, but it is much closer to an actual use-case. I dislike it when an example is unrealistic.

[rBatt]

use the .RData files so that R objects are imported and lazy loaded. Maybe you can do this when reading from text, but I'm not sure. Point is that the package will just load those objects when someone goes to use them, no scripts to actually read in the text.

On the other hand, you have a fair point about RData files – would we let them see how we read .txt in? Seems like this would be hidden anyway. Also, .txt plays nicer w/ GitHub.

[lawinslow]

The lazy loading is only relevant when the data is used within the package. We are just including data as an example and they will never be loaded by the package. Eventually, we will need to pare-down the data we have as our files are too big for CRAN. But I stand by the "real-world use scenario" idea.

[rBatt]

Has this been completed?

[jordansread]

This seems to exist in demo