July 18 2003 3:47:32.668 PM CVT_BASIS Arrange a set of PDE solution data into clusters. Null cluster policy: 0, do nothing, accept null clusters; 1, reset center to a random data point; 2, reset center to random point in hull; NULL_CLUSTER_POLICY = 1 RANDOM_INITIALIZE Initialize RANDOM_NUMBER with user SEED = 12345678 What is the run type? RUN_TYPE = 8 For RUN_TYPE = 8, read in the steady state solution from "STEADY.TXT" and, letting SS be the steady state solution, subtract 5/3 SS from solutions 1 through 250 subtract 1/3 SS from solutions 251 through 500. We DROP the odd numbered solutions. We do NOT normalize the result. What is the name of the XY data file? The file "/a/fs/r1/src-samples/datasets/tcell_flow/xy.txt" contains 4961 lines. X minimum : 0.00000 X maximum : 1.00000 Y minimum : 0.00000 Y maximum : 1.00000 What is the name of the steady state file? Steady state information was read from the file "/a/fs/r1/src-samples/datasets/tcell_flow/steady.txt". Steady max norm = 3.00284 Steady l2 norm = 121.425 What is the name of the first solution file? We believe the number of solution files is 500 The data is stored in an M by N matrix. The "spatial" dimension M is 9922 The number of data points N is 500 All the data has been read into POINT. RUN_TYPE = 8: Thin out the input data points. Thinning increment is 2 Original input data size is 500 Thinned data size is 250 Enter lower and upper number of clusters The program will try to determine the minimum energy of a clustering, for cluster sizes in the range: 4 4 Enter the number of different random cluster configurations to check For each number of clusters, the number of distinct initial random configurations to be checked will be 15 Enter the number of energy iterations For each initial random configuration, the number of times the program will recompute the cluster centers, cluster components, and energy is 30 Enter element file for mass matrix preconditioning or "None". Number of elements = 2400 Number of nodes per element = 6 The bandwidth of the matrix is 165 Compute the mass matrix. Cholesky-factor the mass matrix. Enter 0 to use raw data, 1 to use normalized data. NORMAL = 0 Data will NOT be normalized. Number of clusters allowed: 4 1 Initial_RAW 3.54120 0 HMEANS_RAW 0.840772 7 KMEANS_RAW 0.689109 11 2 Initial_RAW 3.65952 0 HMEANS_RAW 0.691508 13 KMEANS_RAW 0.689109 3 3 Initial_RAW 3.66502 0 HMEANS_RAW 0.692911 12 KMEANS_RAW 0.690447 3 4 Initial_RAW 3.62451 0 HMEANS_RAW 0.691508 12 KMEANS_RAW 0.689109 3 5 Initial_RAW 3.70122 0 HMEANS_RAW 0.690804 16 KMEANS_RAW 0.689109 3 6 Initial_RAW 3.59029 0 HMEANS_RAW 0.840772 7 KMEANS_RAW 0.689109 11 7 Initial_RAW 3.54957 0 HMEANS_RAW 0.692911 10 KMEANS_RAW 0.690447 3 8 Initial_RAW 3.61024 0 HMEANS_RAW 0.692911 12 KMEANS_RAW 0.690447 3 9 Initial_RAW 3.62604 0 HMEANS_RAW 0.692911 11 KMEANS_RAW 0.690447 3 10 Initial_RAW 3.69637 0 HMEANS_RAW 0.692911 13 KMEANS_RAW 0.690447 3 11 Initial_RAW 3.65143 0 HMEANS_RAW 0.691508 13 KMEANS_RAW 0.689109 3 12 Initial_RAW 3.66749 0 HMEANS_RAW 0.690804 16 KMEANS_RAW 0.689109 3 13 Initial_RAW 3.72191 0 HMEANS_RAW 0.691508 14 KMEANS_RAW 0.689109 3 14 Initial_RAW 3.69847 0 HMEANS_RAW 0.692224 16 KMEANS_RAW 0.690447 3 15 Initial_RAW 3.58427 0 HMEANS_RAW 0.840772 7 KMEANS_RAW 0.689109 11 ANALYSIS_RAW: Computed energy range for given cluster size: (The minimum and maximum should be close if we're taking enough iterations.) Number of Minimum Maximum Clusters Energy Energy 4 0.6891 0.6904 Energy table: Number of Energy Clusters Energy /point Sqrt(E/Pt) 4 0.6891 0.0028 0.0525 CLUSTER_CENSUS Individual cluster population and energy Index Population Percentage Energy Percentage Min Max 1 12 4 1575.93 35 126 137 2 17 6 512.222 11 7 23 3 6 2 514.488 11 1 6 4 215 86 1827.08 41 24 250 ------ --- ------------ --- Total 250 100 4429.72 100 1 250 CVT_BASIS: Writing cluster generators to individual files. Write file gen_001.txt Write file gen_002.txt Write file gen_003.txt Write file gen_004.txt CVT_BASIS Normal end of execution. July 18 2003 3:49:11.712 PM