# Clsti

### Purpose

Temperature Interpolated Classical Least Squares models.

### Synopsis

- model = clsti(files); %builds CLSTI model (calibration step)
- model = clsti(files,options); %builds CLSTI model (calibration step)

- Can also build a CLSTI model using the CLSTI Model Builder interface

- pred = clsti(x,temps,model);%makes predictions with a new data and temperatures
- pred = clsti(x,temps,model,options);%makes predictions with a new data and temperatures

### Description

CLSTI (Classical Least Squares - Temperature Interpolated) models consist of a series of pure component spectra collected at a number of temperatures. At model application time, a test spectrum of a mixture of the same pure components and a sample temperature are provided. For each pure component, the spectrum at the supplied temperature is estimated by linear interpolation. Once each of the pure component estimated spectra at the given temperature has been calculated, a CLS model is built with these estimated spectra as reference measurements. This CLS model then applied to the test spectrum (algorithm = least squares). Available outputs from the prediction are:

- scores - these values represent how many multiples of a pure component spectrum are "contained" within the test spectrum; subsequent modification to produce more useful units (for example, weight percent) is typically done
- Q residuals
- Hotelling's T
^{2} - qcon
- error estimates
- the error estimates are based upon equation 5.14 from
*Chemometrics: A Practical Guide*(Beebe, Pell, and Seasholtz, 1998) modified to account for degrees of freedom

- the error estimates are based upon equation 5.14 from

#### Inputs

To build a CLSTI model:

**files**= cell array of definition files, that can be:- .xlsx, .csv, or .txt file formats, with pure component name in first cell and filenames for pure component data in column 1 and corresponding temperatures in column 2, or
- a cell array of DataSet Objects of pure component spectra at different temperatures with corresponding temperatures in the .axisscale{1,1} field.
**Note: The variables used in the model will correspond to the included variables in the first DSO in the cell array.**

To apply a CLSTI model:

**x**= DSO of test data with temperatures in .axisscale{1,1} field**temps**= vector of temperatures (if not in x.axisscale{1,1})**model**= CLSTI model to apply to x

#### Outputs

**model**= standard model structure containing the CLSTI model (See Standard Model Structure).**pred**= prediction object.

### Options

options = a structure array with the following fields:

**plots**: [ {'none'} | 'final' ] governs plotting of results.**display**: [ 'off' | {'on'} ] governs level of display to command window.**blockdetails**: [ 'compact' | {'standard'} | 'all' ] level of detail (predictions, raw residuals, and calibration data) included in the model.

- ‘Standard’ = the predictions and raw residuals for the X-block as well as the X-block itself are not stored in the model to reduce its size in memory. Specifically, these fields in the model object are left empty: 'model.pred{1}', 'model.detail.res{1}', 'model.detail.data{1}'.
- ‘Compact’ = for this function, 'compact' is identical to 'standard'.
- 'All' = keep predictions, raw residuals for both X- & Y-blocks as well as the X- & Y-blocks themselves.