This is a coding/data exercise to be done in class today. Work together on this but only 1 person should be writing the actual code.
This exercise should take 45 minutes or so.
Procedure:
- copy the data filed called cdata.txt on ~dkmatter/sciprog/cdata.txt
- The data is three columns with 121 events. DO NOT SMOOTH THE DATA.
- You will write a procedure to perform numerical integration of the two data sets (column 2 and column 3) in each case over 20 year intervals (i.e. 1900-1920; 1920-1940, etc). The last interval will be 2000-2012.
- The goal here is to determine if there is a statistically significant positive area signal in this data, over these time intervals, or is it all just random noise.
- The statistical error is essentially the square root of the absolute value of your total area. You need to keep track of this statistic per time interval.
- Now numerically integrate over the entire time series and compare that value to the sums of the individual time intervals.
- For column 3 data you will probably want to use "Simpson's Rule" or something like that (this will be discussed prior to assignment.
- For the data in column 3 (which is the smoothed data) write code to detect and identify the year of each local maximum (positive peak) throughout the entire time series to see if there might by any periodicity.
- Upload your area values + errors + local peaks to
http://homework.uoregon.edu/upload/what
|