How do I perform a clustering ?

Step 1 : You must first load your data into ArrayMiner by using one of the following options:

Use a csv file and choose the open file command (show me the required format)
Call ArrayMiner from GeneSpring (tell me more), the data are passed automatically
Use the data stored in the Clipboard (tell me more)
Use the ArrayMiner data import window (tell me more)

Step 2: Edit the run parameters

Choose the algorithm model (tell me more)
Customize the algorithm parameters (tell me more)

Step 3: Wait for the end of loading, check the data in the graphic window, and hit the optimize button. Select the number of clusters you wish to use and begin clustering.

Step 4: Stop the algorithm when an “Optimization Finished” box appears, or at any later time when the displayed solution appears satisfactory, by hitting the stop button. Examine the solution with help of the graphic windows. Resume the optimization from step 3 if appropriate, by hitting again the optimize button.

Step 5: Save your solution or simply return to GeneSpring, depending on the way you loaded your data in step 1.

NOTE

If you want to test different number of clusters, simply add the current solution to the classification compare window at step 4, hit the optimize button and select the new number of clusters. You can revert to a classification in the classification compare window by right-clicking on it and select the "Display in graphical window" option. To save the content of the classification compare window, you have to save your project as an amg file or save each classification as an amc file.