GoldenCheetah metric exports for my master thesis

[natefontaine]

That sounds good, what export would you like? also does it matter which version of GC we have? Maybe include a step by step on exactly what you want from us

[dawe]

That sounds good, what export would you like? also does it matter which version of GC we have? Maybe include a step by step on exactly what you want from us

The version doesn't matter. My goal is to be independent of the used GC version.

In my GC the menu entry can be found at "Activity" -> "Export Metrics as CSV..."
Then just choose a name for the file and store it somewhere. Attach it to a mail addressed to me and that's it.
You would really help me a lot with this

[mentok]

I suggest that you cross post your request to the GC google list. You will find a lot of helpful users there I would think.

[dawe]

I suggest that you cross post your request to the GC google list. You will find a lot of helpful users there I would think.

Thanks for the suggestion. Unfortunately the reponse isn't that big so far

[boysa]

Sorry, I'm not exactly clear what the purpose is. Maybe others feel the same, and that is why we are reluctant to give out our data?

[dawe]

Okay, I will explain in some more detail what I'm doing in my thesis.

First, what is a training model?
Training models are mathematically defined models that try to simulate how the human body adapts to training loads.
These models work with some individual parameters that are derived from collected training data. (That's why I need the metrics exports.)
Then they take a series of training loads as input and return a series of performance numbers as output.

What am I working on?
The system I'm working on currently supports the Fitness Fatigue model (aka Impulse Response) from Banister et. al. and the PerPot model from Perl.
The goal is to evaluate, if these models can be used with real world data not collected in a lab environment and if automatically generated plans based on these models can help cyclists to reach their goals.
Users of the system will be able to say what level of performance they want to reach in how many weeks, what the maximum training load should be and what weekdays can be used for training.
The system will then generate a plan of training loads that hopefully leads exactly to the given goal. Of course this only works as far as the model is able to simulate the body of the user.

Right now, I'm just asking for data to be able to test the system for different athletes. The test data won't be published or disclosed in any way.