time_series, a dataset directory which contains examples of files describing time series.
A time series is simply a sequence of values of a time dependent quantity. Often, the samples are taken at a constant time interval. Examples of time series include the Dow-Jones stock index, the daily noon temperature reading at the Roanoke airport or the yearly estimate of Mexico's population.
Given a time series, there are some natural tasks that might arise, including interpolation (estimating values at intermediate points), integration (estimating the area below the curve from which the samples were taken) or Fourier analysis (searching for underlying "rhythms" or "tides" or natural cycles that compose a significant portion of the observed data).
The information on this web page is distributed under the MIT license.
FFTPACK5, a FORTRAN90 code which implements the Fast Fourier Transform.
FFTW3, C programs which illustrate the use of the FFTW3 library for Fast Fourier Transforms.
NMS, a FORTRAN90 library which includes a wide variety of numerical software, including solvers for linear systems of equations, interpolation of data, numerical quadrature, linear least squares data fitting, the solution of nonlinear equations, ordinary differential equations, optimization and nonlinear least squares, simulation and random numbers, trigonometric approximation and Fast Fourier Transforms.
sftpack, a C++ code which implements the "slow" Fourier transform, intended as a teaching tool and comparison with the fast Fourier transform.
stats, a dataset directory which contains datasets for computational statistics;
air_speed is a table of M = 20 rows and N = 2 columns. Column 1 is the year, column 2 the record air speed in miles per hour. This data was listed by Lienhard.
airline_passengers is a table of M = 144 rows and N = 3 columns. Column 1 is the year, from 1949 to 1960, column 2 is the month, from 1 to 12, and column 3 is the number of international airline passengers in thousands. This data was presented as series G by Box, Jenkins and Reinsel, 1976.
beveridge_wheat is a table of M = 370 rows and N = 2. Column 1 is the year, from 1500 to 1869. Column two is an averaged price for wheat. This data was presented by Beveridge, 1921.
bulgaria_population is a table of M = 19 rows and 2 columns. Column 1 is the year, and column 2 is the population of Bulgaria.
co2 is a table of M = 216 rows and N = 4 columns. Column 1 is a month index which simply runs from 1 to 216; column 2 is the measured level of carbon dioxide; column 3 is the numeric month (1-12), and column 4 is the year of the reading. This data was derived from the book by David Kahaner, Cleve Moler, and Stephen Nash, "Numerical Methods and Software".
ecg is a table of M = 4096 rows and N = 1 columns. Column 1 represents simulated measurements of an electrocardiogram signal, over 8 cycles. The data was generated in MATLAB using the commands x = ecg(512); x = repmat ( x, 1, 8 ); x = x + 0.1 * randn ( 1, length ( x ) );
el_nino is a table of M = 168 and N = 2 columns. Column 1 is a month index which simply runs from 1 to 168; column 2 is the difference in atmospheric pressure between Easter Island and Darwin, Australia, measured at sea level at the same moment each month. This data was derived from the book by David Kahaner, Cleve Moler, and Stephen Nash, "Numerical Methods and Software".
gnp is a table of M = 258 rows and N = 2 columns. Column 1 is the data, presented as a decimal year, from 1947.00 to 2011.25. Column 2 is the quarterly GNP estimate in billions of dollars, not adjusted for inflation. The data comes from the US Department of Commerce, and is explained at http://www.bea.gov/national/pdf/nipaguid.pdf.
hog_price is a table of M = 82 rows and N = 2 columns. Column 1 is the year, from 1867 to 1948, and column 2 is the 1000 times the logarithm of the price per head on January 1. This data was presented as series Q by Box, Jenkins and Reinsel, 1976.
hurricanes is a table of M = 141 rows and N = 2 columns. Column 1 is the year and column 2 is the number of Atlantic hurricanes.
ibm_stock is a table of M = 369 rows and N = 1 column. Column 1 is the closing price of IBM Common Stock for each trading data from 17 May 1961 to 2 November 1962. This data was presented as series B by Box, Jenkins and Reinsel, 1976.
land_speed is a table of M = 16 rows and N = 2 columns. Column 1 is the year, column 2 the record land speed in miles per hour. This data was listed by Lienhard.
lynx is a table of M = 114 rows and N = 2 columns. Column 1 is a year index from 1821 to 1934; column 2 is a count of the number of lynx captured that year in the Mackenzie River district of Canada, to be sold by the Hudson's Bay Company.
measles_nyc is a table of M = 432 rows and N = 2 columns. Column 1 lists months from January 1, 1928 to November 1, 1963, as decimal values. Column 2 lists the number of measles cases reported over that month.
nile is a table of M = 570 rows and N = 1 column. Column 1 is the maximum flow volume of the Nile river for each year.
playfair is a table of M = 50 rows and N = 3 columns. Column 1 is the year, column 2 the price of a measure of wheat in shillings, and column 3 is the average weekly earnings of a mechanic in shillings. The interesting plot is the ratio of wheat price to earnings.
prices is a table of M = 242 rows and N = 13 columns. Column 1 is the month, column 2 the year, and columns 3 through 13 are average monthly prices for 11 consumer products.
snowfall is a table of M = 135 rows and N = 10 columns. Column 1 is the winter year identifier, ranging from 1890-1891 to 2024-2025. Columns 2 through 9 are the snowfall in inches for October, November, December, January, February, March, April and May. Column 10 is the total snowfall. These measurements were taken near Michigan Tech.
state_number is a table of M = 234 rows and N = 2 columns. Column 1 is the year, column 2 is the number of US states.
sunspot is a table of M = 314 rows and N = 2 columns. Column 1 is a year index which runs from 1700 to 2014; column 2 is a count of the number of sunspots observed that year.
us_population_census is a table of M = 23 rows and N = 2 columns. Column 1 is a census year index which runs from 1790 to 2010 by 10's; column 2 is a count of the US population in that year.
us_population_1900s is a table of M = 100 rows and N = 2 columns. Column 1 is a year index which runs from 1900 to 1999; column 2 is an estimate of the US population in that year.
wing_oscillation is a table of M = 329 rows and N = 2 columns. Column 1 is the time, and column 2 is the Y coordinate of the wing tip of a Boeing 737. This data was reported by Rhett Allain, in Wired Magazine, "The physics of why airplane wings oscillate in turbulence".
wometco_sales is a table of M = 39 rows and N = 2 columns. Column 1 is the index of the month and column 2 is the monthly sales data.
make_plots is a MATLAB script that generates some of the plots.