scientific_paper
Scientific paper datum for retrieval and re-reading. A lot of the variables are there for fluff & flavor.
Vars | |
abstract | Abstract. |
---|---|
author | The principal author of our paper. |
et_alia | Whether this paper is co-authored or not. |
experiment_path | Experiment typepath. |
gains | The coop and funding gains from the paper. |
partner_path | The selected sponsor for our paper. Pathtype form. |
tier | Derived from tracked_variable. Used for indexing and to reduce duplicates. Only one paper can be published in each tier for each experiment. |
title | The title of our paper. |
tracked_variable | The main "score" of a particular experiment. |
Procs | |
allowed_to_publish | |
calculate_gains | Calculate the gains of an experiment. Gain calculation follows a sigmoid curve. f(x) = L / (1+e^(-k(x-xo))) L is the upper limit. This should be the gain variable * 2. k is the steepness. x0 is the midpoint. x is our tracked variable. Returns the expected value of that tier. |
calculate_tier | Determine which tier can we publish at. Lower limit for an allowed tier is 10% of gain. Empty list if none are allowed. |
clone_into | Clones into a new paper type. Important (non-fluff) variables will be carried over and should be cleaned with set_experiment by whoever is calling this. |
return_gist | Returns the formatted, readable gist of our paper in a list. |
set_amount | |
set_experiment | Experiment -> Tier -> Gains Experiment -> Partners -> Gains Changing anything in the chain means those following it should be recounted. |
set_partner | |
set_tier | Sets a tier for us. Nulls the tier when called without args. Re-counts the amount. |
Var Details
abstract
Abstract.
author
The principal author of our paper.
et_alia
Whether this paper is co-authored or not.
experiment_path
Experiment typepath.
gains
The coop and funding gains from the paper.
partner_path
The selected sponsor for our paper. Pathtype form.
tier
Derived from tracked_variable. Used for indexing and to reduce duplicates. Only one paper can be published in each tier for each experiment.
title
The title of our paper.
tracked_variable
The main "score" of a particular experiment.
Proc Details
allowed_to_publish
- Fully check if our paper have all the required variables, and prevent duplicate papers being published in the same tier.
- Things to check: tier, gain, and partner here. ex_path and record datums in subtypes.
calculate_gains
Calculate the gains of an experiment. Gain calculation follows a sigmoid curve. f(x) = L / (1+e^(-k(x-xo))) L is the upper limit. This should be the gain variable * 2. k is the steepness. x0 is the midpoint. x is our tracked variable. Returns the expected value of that tier.
calculate_tier
Determine which tier can we publish at. Lower limit for an allowed tier is 10% of gain. Empty list if none are allowed.
clone_into
Clones into a new paper type. Important (non-fluff) variables will be carried over and should be cleaned with set_experiment by whoever is calling this.
clone_into your own typepath will be like a normal clone.
If you want to subtype this, do it in a way that doesn't mess with the type change.
return_gist
Returns the formatted, readable gist of our paper in a list.
set_amount
- Does the counting for gains.
- Call this whenever experiments/tier/partners are changed.
- Handled automatically, calling this proc externally shouldn't be necessary.
- Also doubles as an initialization for the gains list.
set_experiment
Experiment -> Tier -> Gains Experiment -> Partners -> Gains Changing anything in the chain means those following it should be recounted.
Used when assigning an experiment to a specific paper. Failing to provide a proper path, a tracked variable, or a correct data should null every non-fluff data. Implement this in the children procs.
set_partner
- Sets a specific partner for our paper. Partners give money and boost points.
- Partners exist separately from experiments and wont need to be reset every time something changes.
set_tier
Sets a tier for us. Nulls the tier when called without args. Re-counts the amount.