2

I have a telemetry data which consist of its position and the activity of a bird. The dataset is in the csv format which I am uploading:

Lat. Long. Act    Date   Time
12    17   Eat    5-1-08 13:10
14    18   Rest   5-1-08 13:30
19    14   Walk   5-1-08 13:50
11    10   Rest   5-1-08 14:10
18    11   Walk   5-1-08 14:20
15    19   Walk   5-1-08 14:40
10    17   Rest   5-1-08 15:00    
....

I have produced ggplot (lat and long) to get the clusters over the bird path. Theory is higher the density of a cluster (in a continuous time interval) probability that it is eating is more.

plt1 <- ggplot(data = dataframe, aes(y = Lat., x = Long., colour = Act) )

Problem:

  1. I have to find patterns (& store it) what the bird is doing after eating and for some unknown dataset (only Lat. and long is provided) I have to find the eating points in the plot based on the known patterns.

Can anyone help me in suggesting some solution or references (algorithms, libraries.. anything..) to this..?

SabDeM
  • 7,050
  • 2
  • 25
  • 38
  • This sounds more like a methods question for CrossValidated than a programming question for StackOverflow. – ulfelder Jul 26 '15 at 11:04
  • @ulfelder It is a programming question. –  Jul 26 '15 at 11:08
  • 1
    I guess I see it as a methods question if you're not sure what modeling or pattern-recognition function to call to accomplish the task, and a programming question if you know the function you want but are not sure how to execute it. Anyway... – ulfelder Jul 26 '15 at 11:10
  • 1
    In this, I want to fist make my data frame continuous let's say in 30 minutes time interval. Then I want to find the dense clusters. Also, I have to find the eating pattern of a bird store it. Then based on that I want to predict for the "Eat" points where only lat and longs are given. –  Jul 26 '15 at 11:12
  • @ulfelder yes you can say that. I am confused how to tackle the problem –  Jul 26 '15 at 11:13
  • I would recommend looking on CrossValidated for questions about spatial pattern recognition and posting a new one there if you can't find any good leads. – ulfelder Jul 26 '15 at 11:16
  • @ulfelder I'll try that.. thank you. Can you help me out in this problem? –  Jul 26 '15 at 11:17
  • No, I don't know those methods. – ulfelder Jul 26 '15 at 11:17
  • cross-post: http://stats.stackexchange.com/q/163223/7828 – Has QUIT--Anony-Mousse Jul 26 '15 at 18:51

0 Answers0