# How To Do A Basic Analysis

This quick-start guide will give you a very brief script to follow for quickly importing data from text files and doing a basic MAR-1 analysis (with nothing fancy).

1. Start up LAMBDA by opening LAMBDA.exe. Wait for the start page (looks like this). (If you are not using the executable but are using MatLab and the source code, you start by typing 'lambda' at the MatLab cmd line).

2. From the LAMBDA start page, click on the File menu and select Import -> From text file.

3. Navigate through the directory structure until you find your text file and select it.

4. Select which variables in your dataset are Classification variables (year, sample #, etc).

5. Next, select which ones are variates (i.e., species of interest). Whatever is left over (if anything) will be classified as a co-variate.

6. Make sure your data were imported correctly by clicking on the Data menu and choosing the view -> all option.

7. Click on the statistics menu and choose transform -> natural logarithm.

8. Apply any other transformations (e.g., z-score) that you wish using the statistics -> transform option.

9. Click on the MAR-1 menu, select CLS Parameter Estimation, and choose 'CLS Search over state spaceā¦'. You'll be asked some questions. You can accept the defaults for a basic analysis.

10. Once the search is complete, again select MAR-1 -> CLS Parameter Estimation but this time choose 'Run MAR-1 model with CLS estimates'.

11. Your MAR-1 model results will be displayed in the command window. If you are running the executable version, the command window is that black window that appears first when LAMBDA is started.

12. To compute diagnostics, click on the MAR-1 menu and choose Assess model fit (result will appear in command window) and then click on the MAR-1 menu and choose Normal probability plot of the residuals.

13. Click on the MAR-1 menu and choose Compute Information Criteria (result appears in command window).

14. Click on the MAR-1 menu and choose Compute Stability Properties (result appears in command window).

15. To save your results, click on the File menu, and choose MAR -> Save MAR Results.

1. Start up LAMBDA by opening LAMBDA.exe. Wait for the start page (looks like this). (If you are not using the executable but are using MatLab and the source code, you start by typing 'lambda' at the MatLab cmd line).

2. From the LAMBDA start page, click on the File menu and select Import -> From text file.

3. Navigate through the directory structure until you find your text file and select it.

4. Select which variables in your dataset are Classification variables (year, sample #, etc).

5. Next, select which ones are variates (i.e., species of interest). Whatever is left over (if anything) will be classified as a co-variate.

6. Make sure your data were imported correctly by clicking on the Data menu and choosing the view -> all option.

7. Click on the statistics menu and choose transform -> natural logarithm.

8. Apply any other transformations (e.g., z-score) that you wish using the statistics -> transform option.

9. Click on the MAR-1 menu, select CLS Parameter Estimation, and choose 'CLS Search over state spaceā¦'. You'll be asked some questions. You can accept the defaults for a basic analysis.

10. Once the search is complete, again select MAR-1 -> CLS Parameter Estimation but this time choose 'Run MAR-1 model with CLS estimates'.

11. Your MAR-1 model results will be displayed in the command window. If you are running the executable version, the command window is that black window that appears first when LAMBDA is started.

12. To compute diagnostics, click on the MAR-1 menu and choose Assess model fit (result will appear in command window) and then click on the MAR-1 menu and choose Normal probability plot of the residuals.

13. Click on the MAR-1 menu and choose Compute Information Criteria (result appears in command window).

14. Click on the MAR-1 menu and choose Compute Stability Properties (result appears in command window).

15. To save your results, click on the File menu, and choose MAR -> Save MAR Results.

## 1 Comment