Timeseries Tools

Fast, optimized tools for working with timeseries data. Calculates weighted frequency power spectra of un-evenly sampled data, filter and extracts frequencies. Written in Fortran 90 using OpenMP for parallelisation.

Written by Dr. Rasmus Handberg, Department of Physics and Astronomy, Aarhus University.

Installation instructions:

Using existing installation at SAC, Aarhus University

If you are and have access to the computer systems here, the easiest way to use the programs is simply to use the pre-compiled binaries available on the systems.

/usr/users/f041080/tidsserie/bin/machine

where machine is the system you are running on.

If you wish the program to use all available CPU cores available on your machine (and why wouldn't you?), you need to edit the file ~/.bashrc or ~/.cshrc (depending on which shell you use) and add the following line:

setenv OMP_NUM_THREADS 8 (csh, tcsh)
export OMP_NUM_THREADS=8 (bash)

The entire setup for the computer "ast", with 8 CPU cores, would for example look like this (in bash):

export PATH=$PATH:/usr/users/f041080/tidsserie/bin/ast
export OMP_NUM_THREADS=8

If you have any questions, feel free to ask me.

Matlab

setenv MATLABPATH $MATLABPATH:path-to-program/matlab

Python

setenv PYTHON_PATH $PYTHON_PATH:path-to-program/python

Compiling it yourself

1. You should then be able to check out the code from the repo. using the following command:

git clone https://github.com/rhandberg/timeseries.git .

2. Run make followed by make install. This will compile the program (by default using gfortran) and put the compiled binaries into the bin directory. If you wish to use a different compiler, you can simply modify the file makefile.preample where you can specify which compiler (and settings) to use.

3. Add the program to your path like described above.

TODO

• Use Numerical Recipies trick in CalcAlphaBeta to speed up calculation.
• Make automatic or semi-automatic mode for windowfunction.
• Make oversampling-factor commandline input for spec and window.
• Finish writing install instructions.

Documentation

Input files should be ASCII files with three columns: time, measurement and statistical error.

Weighted Least Squares Spectrum (spec)

Input:
  spec [options] [inputfile] [outputfile]

Options:
  -p         Power spectrum (default)
  -a         Amplitude spectrum
  -pd        Power density spectrum
  -tsec      Treat times as seconds
  -tday      Treat times as days (default)
  -kplrraw   Input file is a Kepler data file (use raw columns)
  -kplrpdc   Input file is a Kepler data file (use PDC columns)
  -kplrwg    Input file is a Kepler data file (use WG columns)
  -auto      Automatic. No interaction required (experimental)
  -quiet     Print nothing to screen
  -version   Print program version information

Written by
  Rasmus Handberg
  Department of Physics and Astronomy
  Aarhus University

Spectral window function (window)

WINDOW
Calculate the Spectral window-function
of a timeseries.

Input:
   window [options] [inputfile] [outputfile]

Options:
   -tday      Treat times as days (default)
   -tsec      Treat times as seconds
   -kplrraw   Input file is a Kepler data file (use raw columns)
   -kplrpdc   Input file is a Kepler data file (use PDC columns)
   -kplrwg    Input file is a Kepler data file (use WG columns)
   -quiet     Print nothing to screen
   -version   Print program version information

Written by
  Rasmus Handberg
  Department of Physics and Astronomy
  Aarhus University

Iterative Sine Wawe Fitting (clean)

CLEAN
Iterative Sine Wawe Fitting

Input:
   clean [options] [inputfile] [outputfile]

Options:
   -tday      Treat times as days (default)
   -tsec      Treat times as seconds
   -version   Print program version information

Written by
  Rasmus Handberg
  Department of Physics and Astronomy
  Aarhus University

Iterative Sine Wawe Fitting with limit (lclean)

LCLEAN
Iterative Sine Wawe Fitting

Description:
  Cleans a timeseries until given limit.

Input:
  lclean [options] [inputfile] [outputfile]

Options:
  -p         Power spectrum (default)
  -a         Amplitude spectrum
  -pd        Power density spectrum
  -tday      Treat times as days (default)
  -tsec      Treat times as seconds
  -version   Print program version information

Written by
  Rasmus Handberg
  Department of Physics and Astronomy
  Aarhus University

Simultaneous Iterative Sine Wave Fitting (sclean)

SCLEAN
Simultaneous Iterative Sine Wave Fitting

Description:
  Cleans a timeseries for coherent signals,
  taking into account the influences of the
  frequencies on each other.

Input:
  sclean [options] [inputfile] [outputfile]

Options:
  -tday      Treat times as days (default).
  -tsec      Treat times as seconds.
  -version   Print program version information.

Written by
  Rasmus Handberg
  Department of Physics and Astronomy
  Aarhus University

Lowpass filtering using Weighted Least Squares (lowpass)

LOWPASS
Lowpass filtering using Weighted Least Squares

Input:
  lowpass [options] [inputfile] [outputfile]

Options:
  -tday      Treat times as days (default).
  -tsec      Treat times as seconds.
  -version   Print program version information.

Written by
  Rasmus Handberg
  Department of Physics and Astronomy
  Aarhus University

Highpass filtering using Weighted Least Squares (highpass)

HIGHPASS
Highpass filtering using Weighted Least Squares

Input:
  highpass [options] [inputfile] [outputfile]

Options:
  -tday      Treat times as days (default)
  -tsec      Treat times as seconds
  -version   Print program version information

Written by
  Rasmus Handberg
  Department of Physics and Astronomy
  Aarhus University

Bandpass filtering using Weighted Least Squares (bandpass)

BANDPASS
Bandpass filtering using Weighted Least Squares

Input:
  bandpass [options] [inputfile] [outputfile]

Options:
  -tday      Treat times as days (default).
  -tsec      Treat times as seconds.
  -version   Print program version information.

Written by
  Rasmus Handberg
  Department of Physics and Astronomy
  Aarhus University

Bandstop filtering using Weighted Least Squares (bandstop)