This page provides a description and an example of the files needed to calibrate particle-size distributions measured by different analytical techniques.

Data calibration converts from one data format to another, from FORMAT X to FORMAT Y. The codes X and Y are used because the calibration process uses a model to predict FORMAT Y using data measured by FORMAT X.

What do I need?

Calibration requires a set of duplicate particle-size distributions for model training: sediment samples that have been measured by both FORMAT X and FORMAT Y.

These data should be given in pairs of csv files: one for FORMAT X and one for FORMAT Y.

These files should follow these format restrictions:
  • Each file contains a header row giving the upper limit of each particle-size category in phi[1] .
  • Measured particle-size distributions are given as rows of percentage frequency data.
  • Measurements in both files should be arranged in the same order so that rows in 'FORMAT X' corresponds with the same rows in file 'FORMAT Y' .
  • Both files should therefore have the same number of rows.
  • A minimum of thirty samples measured by each technique is recommended. For a highly variable data more samples are recommended.

TECH_CODE: Some data files contain a mix of different techniques. To convert from one technique to another each particle-size record should be labelled with a specific technique code. See TECH_CD in Input file formats.

Download an Excel version of this template here. Each worksheet can be saved as a 'csv' file from Microsoft Excel or Open Office.

IMPORTANT: The same code in TECH_CD will be used to label FORMAT X and FORMAT Y when you create a calibration model. The codes must match for calibration to work. TNO users should use the codes given here for DINO, DINO_Qwa and TOP_INTEGRAAL files..

The function used for data calibration is called dataCalibrate. Enter doc dataCalibrate into the Matlab command window for more information.
  1. ^ To convert from micrometers to phi use the microns2phi function or refer to this conversion table: