Data management - how do you do yours?

[Tim Langlois]

Currently we have a lot of data from different projects in an Access database  - that has queries to calculate biomass.

But I am interested in developing workflows in R to take data direct from Event measure and through to analysis and plotting.

Also keeping an eye on the development of on line databases that may store data at some point in the above workflow. 

Tim Langlois - timothy....@uwa.edu.au

[Dom Andradi-Brown]

Hi Tim,

I know this post is a bit old now, but I'm just starting to think about writing some functions for R to take a concatenated csv file that EventMeasure outputs and plot data for requested species/families as I'm going to be doing a lot of that over the next couple of years.

Previously I've been working for Operation Wallacea, and their stereo-video data is basically on a series of spreadsheets for each of their field sites and years (we currently only do diver operated stereo-video), and analysis hasn't really moved into R because a lot of UG students use the data for their projects so they want to be able to easily share chunks of data on spreadsheets. But I'm excited about the idea of building some functions allowing quick and easy data exploration and analysis directly from the outputs of EventMeasure.

If you've started developing functions or want to collaborate to write some functions for R I'd be very interested!

Cheers,

Dom

[timothy....@uwa.edu.au]

Hi Dom

Yes - great idea.

I have been working a lot recently on scripts to take the data straight from eventmeasure output files.

Part of this is trying to keep integrety between the video adn .emobs files and the data that get used in analyses.

So that if you go back and make any changes in the .emobs - you should be able to simply run the script again to produce plots/analyses!

Let me know exactly what you are keen on doing and I will see if I have scripts to fit.

We should prob keep posting them for other to see - would be good to write up standard scripts.

You will see the data checking script I put up here before - that is useful - but is now i think implemented in the new version of eventmeasure - not that I have used it yet. 

Cheers

Tim

[timothy....@uwa.edu.au]

Hi Dom

In reply to you message:

Thanks, at the moment I'm looking to write a script to turn fish lengths into rough biomass estimates for a lot of transects that were swum with stereo-video at different sites with different fishing pressures/history.

I have some local length to weight data from a fisheries monitoring program that I want to use, but a fair number of species on the transects don't have any local data, so was thinking of using the fishbase R package so that I can automatically pull data from fishbase for any species without local data. I also have some fish on the transects that are only marked as points as we couldn't get accurate lengths/they only appeared in one camera, so I want to be able to assign them mean (or perhaps median, need to have a think about length distributions) lengths for that species from other individuals of that species I've surveyed.

Sounds like a great idea. I didn't know about the fishbase R package. That is great you have the 3D points - I haven't used them to assign a mean length/biomass yet - but I want to get this scripted. Steve Linfield has been doing this is access.

As this is to support the OpWall coral reef monitoring program ideally I want to wrap it all up nicely in a couple of R functions/scripts so that other people can use it to generate biomass estimates of different fish families etc.

 Great idea

I'm going to start to try and put this together, but if you have written any bits of this pipeline that I could use that would be great and I'll write it so that everything fits together and I'll post the finished code once I've got it working.

 I will email you some examples of scripts that I've been using for calculating biomass, when can improve it by sending it backwards and forwards between us - then re-post on the group

Thanks,

Dom