2 Getting started

 2.1 Displaying a time series
 2.2 Basic control of the plot window
 2.3 Meta data
 2.4 Making a periodogram
 2.5 Folding time series data
 2.6 Fitting folded data

The following sections describes how to display, and carry out basic analysis of, a time series. If you’re new to FROG then do take the time to read this through and try out the example commands.

2.1 Displaying a time series

To start FROG you should just need to type the command:

    % frog &

This assumes that you have FROG installed on your system as part of a standard Starlink installation. If not then you’ll need follow any installation and pre-startup instructions that you have before using this command.

When FROG appears it should look something like:

This is the main browser window. This window acts as a container for your time series analysis task (in what has traditionally been referened to as a ‘stack’

To open your first time series select the Open Time Series item in the File menu. This will create an open file dialog window. Just navigate to your spectrum, select it, and press Open to proceed.

Alternatively you could have supplied the time series you want to use on the command-line:

    % frog <time series> &

Which will read the time series stored in the named 10, see section 10.

If all goes well, your time series will be displayed in a plot window.

The window is labelled Time Series 1. Each time series you read will have a unique identifier (equivalent to a ‘slot’ in command line applications like PERIOD ). In this example we have read in a file consisting of x, y and y-error. The data points are shown in blue, with the error bars in red.

2.2 Basic control of the plot window

The plot window has four menus: File, Display, Graphics and Operations.

You can turn off the display of error bars for this plot using the Display=>Error Bars toggle. If your data is in magnitudes you can flip the y-axis display so that your data is displayed the correct way up using the Display=>Magnitudes toggle. This toggle will propogate through any analysis you do to the data, such as detrending, fitting or folding.

Additionally the Display menu gives us basic control over plot window, allowing us to change the line style and how the plot is scaled in the window.

2.3 Meta data

The File menu offers several options, the two you need to know about immediately are File=>Meta Data and File=>View Data.

The File=>Meta Data item allows you to inspect the information known about your data

this information is known as “meta data”, and is “data about (describing) your data”. Initially little will be known about the data, but more meta-data will accumulate as you perform operations on the data.

The File=>View Data item will pass your data to the Starlink TOPCAT application where you can carry out sophisticated selections, cuts and and analysis on your time series data.

At the moment the only way to reimport you data into FROG is to save it as a FITS format file and load it back into the application using the File=>Open Time Series item. In the near future TOPCAT should offer the option to push your data back to a running instance of FROG using 9, see section 9.

2.4 Making a periodogram

The Operations menu offers you several options, lets first look at the Operations=>Find Periodicities item. Selecting this will generate a Periodogram Analysis popup window.

This popup will try to suggest reasonable values for the minimum and maximum frequency, and frequency interval, based on the Nyquist frequency of the data itself. Alot of the time the suggested values are reasonable, but you should use your common sense, the larger the number of “frequency steps” the longer the periodogram will take to generate.

Currently there are two type of periodogram analysis available, Fourier Power Spectrum and Chiq-squared Periodogram. For now lets go with the default options. Hit the Ok button.

This will generate a fourier transform of your data, labelled periodogram 1 and an associated meta-data popup.

Meta-data popups are automatically generated every time you carry out an operation on your data to keep you informed. You can turn off the appearance of these poups by using the toggle swicthes in the Options=>Automatic Display menu.

2.5 Folding time series data

From our newly created periodogram window we can find the best fit period and fold our original data around this value. In the periodogram window select the Operations=>Find Best Period item. This will automatically find and mark (with a red line) the best period in your Fourier Transform.

You can now select the newly activated Operations=>Fold around Best Period item in the periodogram window, aloowing you to either fold the data, or fold and bin-up the data. The best fit period will be inserted by FROG in the Fold Time Series dialog, along with an arbitarily chosen zero point, and (if necessary) a suggested number of bins.

The folded time series will be labelled Time Series 2. Inspecting the meta-data for the series shows that it derived from Time Series 1, our orginal series.

2.6 Fitting folded data

Selecting the Operations=>Fit Data item from the Time Series 2’s plot window fits the data with a sin() + cos() function.

Here (above) we can see our original data folded around the best fit period found in our periodogram fitted with our sin() + cos() function.

Inspecting the meta-data, using the File=>Meta data item

shows the fit parameters in more detail.

We’ve now walked through the basic proceedure for reading in a time series, finding the best period, and the folding and fitting around this period. However FROG has many other capabilities, so keep reading...