Due Monday June 6: Find a time series data set of at least 100
points from your field and create an ASCII file that TIMESLAB/R
can read. Use TIMESLAB/R to read and plot the data (make sure to use the
option on the command to
put a good caption on the plot, i.e., a good title for the plot.
Use Microsoft WORD or
WORKS to create a brief report that 1) describes the data and what you
would like to learn from it, and 2) imports the graph into the
report.
Due Monday June 13:
Problems C1.17, T1.2, T1.12, T1.14 in the text.
Using power transforms and differencing if necessary, try to
stabilize the variance of your data set and remove trends and cycles
(these might not be necessary). Then use the wntest function to
test it for white noise. Write a report describing your analysis.
Use the ``Time Series/Log Spectra'' applet until you get all the
answers correct four times in a row.
Due Thursday June 23:
Problems T2.2, T2.5, T2.7, T2.13 in the text.
Due Thursday July 7:
Problems C2.5 (use the ARCORR function), C2.7 (write an R function),
C2.9 (use the R function called arma).
Due Thursday July 14:
Use the ARMA(p,q) tester applet until you get at least 10 correct
identifications.