FitsLibrary

This project is currently a very early WIP, use at own risk! There is no nuget package released until the first "release" version is ready

FitsLibrary is a native C# / dotnet core implementation using the most up to date FITS 4.0 specification for writing and reading astronomical FITS files.

The library focuses on being fast and easy to use. It makes use of the newer C# generic maths features, which require net7.0.

What currently works

• Loading of header data
• Validation of header content
• Reading of N-Dimensional data arrays

What doesn't work

• Reading of extension headers
• Writing .fits files

Usage

Open a fits file using

var reader = new FitsDocumentReader<float>();
var fitsFile = await reader.ReadAsync("SampleFiles/FOCx38i0101t_c0f.fits");

You could then iterate over a file with 2D using

for (int x = 0; x < document.Header.AxisSizes[0]; x++) {
    for (int y = 0; y < document.Header.AxisSizes[1]; y++) {
        var valueAtXY = document.GetValueAt(x, y);
    }
 }

Note that the generic parameter (float in the above example) has to match the datatype in the fits file.

If you do not know the datatype of the fits file beforehand, you can use FitsDocumentHelper to get the type of the data inside the file:

var dataType = await FitsDocumentHelper.GetDocumentContentType("SampleFiles/FOCx38i0101t_c0f.fits");

The FitsDocument can take any I/O Stream and work with it.

Accessing Header values

Header values can have different data types (string, integer, float, ...)

They can be read using

var headerValue = fitsFile.Header["TestHeaderKey"] as string

Tip:

If you only want to load the header data of a file you can use below code instead, which will be much faster and only read and parse a files header:

var header = await FitsDocumentHelper.ReadHeaderAsync("SampleFiles/FOCx38i0101t_c0f.fits");

Accessing Content data

Data can be accessed in different ways:

By coordinates

To get the value at specific coordinates, do

var valueAtXY = fitsFile.GetValueAt(x, y);

RawData

By accesing fitsFile.Content (which is of type Memory<T>, use .Span to access data or much slower .ToArray()) This is used for fast access, does not differentiate between dimensions. Index for value in 2 dimensional data for example is calculated like this:

index = indexAxis1 + (axisSize1 * indexAxis2)
fitsFile.Content.Span[index];

There is a typed functions for all supported data types by the fits standard (byte, 16-bit integer, 32-bit integer, 64-bit integer, 32-bit float, 64-bit float)

Be aware, these methods make no sanity checks for performance reasons. If you enter coordinates that do not exists or exceed their length expect random numbers to be returned! Also, if you try to access data in the wrong datatybe, like for example as int32 while data is int64, expect exceptions. To check the datatype of the document use fitsFile.Header.DataContentType.

The sizes of the axis can be determined using:

var numberOfAxis = fitsFile.Header.NumberOfAxisInMainContent;
var axis1Size = fitsFile.Header["NAXIS1"] as long;
var axi2Size = fitsFile.Header["NAXIS2"] as long;
...