search

forc-db / Global_Productivity

Creative Commons Attribution 4.0 International
2 stars 0 forks source link

change code to use |latitude| (not latitude) as fixed effect #6

Closed teixeirak closed 5 years ago

teixeirak commented 5 years ago

@ValentineHerr,

A quick but important fix in your code-- please use the absolute value of latitude, not raw latitude, in this analysis.

ValentineHerr commented 5 years ago

I just pushed the updated version.

It seems that we had made some important changes in the analysis (at least the plotting) that I had never pushed (maybe it was a work in progress? I can't remember). So there are big changes in this last commit. If that is too different to what we had and you don't like the new version, please revert the last changes and we can discuss what to do.

ValentineHerr commented 5 years ago

@teixeirak, I just realized that that last push was not using the latest version of ForC_simplified. Before I push the newer version I'll wait for you to tell me so, just to make it easier to cancel the last commit if we don't like it.

teixeirak commented 5 years ago

It looks like your changes are generally good, although I'm confused by the age interaction plots. I don't think we want to revert that commit.

As you're probably aware @beckybanbury is now making rapid progress on data and planning analysis (issue #5) for this project.

ValentineHerr commented 5 years ago

Ok.

I pushed new results using latest ForC_simplified.

I agree that age interactions plots are confusing. It took me a while to understand what I did back then. For each variable, there is 3 lines, one for age 50 (thinest), 150 and 500 (thickest). If the lines are parallel it means it is just an additive model and if they are not (and there is "interaction" in the equation) it means the interaction of age was significant.

teixeirak commented 5 years ago

@beckybanbury, note that we have some interesting results here to review.