danovando / global-fishery-potential Goto Github PK

• 94: check calculation of FvFmsy
• 96: change c_num
• 98: change c_den

• change names
• 195: change c_num
• 197: change d_den
• 214: change MsyProfits
• 230s: change status quo policies where needed

Line 23: change to reflect new relationships between FvFmsy BvBmsy and 2

• line 18: replace ri with gi
• line 26: change column names to g instead of r
• line 44: replace scahefer growth function with PT

Write code that takes your standardized data table and makes plots of catch, biomass, etc. by size groups and by region

Figure out what is so slow in FormatForRegression

Use the new dataframe to calculate the standard deviations of the residuals for each bin in each year of each model

• 486: change relationship among K and B/Bsy
• 488: change relationship between g and MSY etc.
  *490: change names
• 535: change biomass calculation

Make a plot of proportional error against BvBmsy, color coded by catch size instead. Seems like a clearer relationship

Your life is going to be a lot easier this way, rather than chasing dataframes around the place. Put them all together and make sure that you don't have redundant names

You're not plotting the timeseries of B/Bmsy right from AnalyzeFisheries. You're leaving out the CatchMSY fisheries and their associated errors. Doesn't affect anything else but it's a problem.

Add in a column to DatabaseBuild that provides phi for each fishery. For now set at 1.118, following Branch et al. 2013

• 77 Change profit function
• 79 change bnext

Rewrite the good old lumping code, same as synthetic stocks code

Changes in the actual body of MatrixSnowCatchMSY

• replace all r with g
• 220: change relationship between MSY, r, and K
• 220: change relationship between MSY, r, and K
• 223: same as above
• 250: MSY derivation
• 252: BvBmsy and K relationship
• 256: check relationship between FvFmsy, MSY , catch, and BvBmsy
• 267: more MSY problems
• 268: Fmsy relationship
  *302 Check relationship between FvFmsy, MSY , catch, and BvBmsy

How do we deal with the fact that RAM allows B/Bmsy above 2, but CatchMSY doesn't? The NEIs are inherently different now since they are drawing from RAM, while the FAO's aren't.

Make a much cleaner calculation of country and fishery NPVs etc. using ddply and mutate

Run projection Analyses you did in matlab in R... pain my ass

Take Chris' matlab code for dynamic optimization and get it up and running in R

using the function FindFishBase, take the FullData dataframe, and the ideas in runLinkedPRM, and produce a new version of FullData with missing life history variables filled in where possible

Go back and add species category as a selectable input variable

Write a function that takes some options and cleans up the database before it gets passed to regression (omit NA's in catch, interpolate catch, etc.)

write a basic function to make kobe plots, you'll have to do them a lot

There's a problem in RunProjections where historic RAM FvFmsy values are being overwritten. You need to fix that once you decide how to deal with the mixed values

The ResultMatrixTable starting on line 812 of the wrapper have some hard coded columns in there. Can you check and make sure that these are referencing the right columns, and change these to an automated format?

Thanks

ResultMetricsSQCumFinalTable[c,1]<-CountriesToRun[c]

ResultMetricsBaselineTable[c,2:13]<-FinalYear[FinalYear$Policy=="CatchShare",13:24]

ResultMetricsSQFinalTable[c,2:4]<-FinalYear[FinalYear$Policy=="CatchShare",28:30]

ResultMetricsSQFinalTable[c,5:7]<-FinalYear[FinalYear$Policy=="CatchShare",25:27]

ResultMetricsSQFinalTable[c,8:10]<-FinalYear[FinalYear$Policy=="CatchShare",34:36]

ResultMetricsSQFinalTable[c,11:13]<-FinalYear[FinalYear$Policy=="CatchShare",31:33]

ResultMetricsSQCumFinalTable[c,2:14]<-Cumulatives[Cumulatives$Policy=="CatchShare",2:dim(Cumulatives)[2]]

Not wanting to modify the Wrapper script, I made notes next to code lines that should now be unnecessary. code and notes should be deleted.

Change the parts of DatabaseBuild that calculate reference points for RAM stocks based currently on Schaefer dynamics

In the last paper we assigned them to the nearest present species code. Do that again?

• Change names

• check line 103

Lines that need attention

1. 21
2. 23
3. 27
4. 47
5. 45
6. 49
7. 115

OK, suddenly we aren't getting projections for any NEI stock. What's happening is that they are getting infinite values for BvBmsy in NearestNeighborNeis. I checked in the results of AnalyzeFisheries, and the only infinite values are in NEI or unidentified stocks. My only hunch, maybe it's trying to pull from NEIs for compstocks? If you try and take any function of an infinite you get an infinite, so maybe it's tryign to take like median BvBmsy of an NEI stock which is somehow in the compstocks?

The other possibility, any number dicided by 0 is infinite. So, maybe when you're calculating the denominator (the Bmsys), there's some situation that would result in Bmsy of 0, resulting in an infinite BvBmsy?

Testing

Remember what you needed to store for retransformation bias and store it

For some reason the biomass values are coming up as characters, fix please!