search

stat157 / background

0 stars 4 forks source link

Error Analyzer's Task Division #42

Open bonghyun5 opened 10 years ago

bonghyun5 commented 10 years ago

Overall Goal: Refactor the code to generate an error diagram given a dataset and a model.

Code is given: https://github.com/aparker92/error_analyzers/blob/master/etas-training.R

Each member of the group will try to create a function that will output what we need for the overall function to generate the plot.

Laura: Find out how etas.CI is given to us. Look at lines [25 - 27] in above R code.

Alyssa: Coordinate with the analyzer group 2 to homogenize the output for MDA model with the output for ETAS model we have now.

Tay: Given the earthquake data, divide them up to {times, magnitude}. Given start time and end time, get the period for them. [Lines 1 - 20]

Alice: Given you have training period, get number of training and testing set. Generate place holders for CI.dist and CI.list [Lines 29 - 57] Given you have parameters for ETAS model, and given you have the ETAS model, fill in CI.dist. [Lines 24-27] [Lines 57-87]

Disi: Given you have parameters for ETAS model and given you have ETAS model, fill in CI.list, which is values at event time. [Lines 89 - 112]

Bonghyun: Compile Everything from above Given you have CI.list and CI.dist, generate sorted list for error plot [Lines 116-135] Given error plot numbers, plot it to error plot. [Lines 140-144]

John: Given list of timeperiod, and the Error Points, make a function to get the Area Under the Curve.

DEADLINES: Everyone Except Bonghyun: 11/24 midnight. Bonghyun: Monday night to compile everything.

GalaxyNight-day commented 10 years ago

Check your email !

bonghyun5 commented 10 years ago

So far, only tay has given me his part. Can you guys try to finish by tonight so i can compile everything together?

johnrisko commented 10 years ago

Hey could we go over this in class? My problem is I need a function to implement the integrate function. The plots are a list of data. We need to first find a function that fits the observed data. If possible, can I work on this with someone more technical because it will require fitting the data to a distribution or a function.

On Mon, Nov 25, 2013 at 11:22 PM, Bonghyun Kim notifications@github.comwrote:

So far, only tay has given me his part. Can you guys try to finish by tonight so i can compile everything together?

— Reply to this email directly or view it on GitHubhttps://github.com/stat157/background/issues/42#issuecomment-29271989 .