ILNumerics^® Data I/O

ILNumerics allows reading and writing of array data to/from a number of formats. In the following an overview of the different formats is provided. To learn more about a specific one, just follow the indicated link.

Text / CSV

By applying csvread<T>() and csvwrite<T>(), you can read from and write to CSV files. Furthermore, it is possible to read from custom formats. Find out more here.

Image Data

The following functions are provided to facilitate the efficient loading of image data:

loadImage loads all channels of an image as pixel data into Int32 elements.

loadChannels loads all channels of an image as pixel data into individual matrix slices.

Matlab

ILNumerics.MatFile5 is able to create *.mat files version 6, which are compatible to any common *.mat file format reader/ writer, including Matlab^®. Newer versions of Matlab store data in HDF5 format. Here, you can learn more about generating .mat files.

HDF5

ILNumerics provides a full featured HDF5 API. It brings a convenient object oriented interface and efficient integration into the ILNumerics memory management. Visit the HDF5 documentation here.

Element Pointers & Span

When it comes to low-level element access and if the CLR host /your CPU is sufficient, ILNumerics arrays allow to read and write to their values directly in very efficient ways. Read more.

XML, Databases, Excel Spreadsheets, Native Libs, ...

Based on the rich .NET ecosystem, ILNumerics allows the retrieval and storage of array data in any format accessible by .NET - easily and out of the box. ILNumerics Array allows the import and export of data as well as the efficient access of internal array storage.

Read more:

• Importing and exporting array data,
• Casting arrays from / to system types,
• Low-level element access (pointers, Span,...)