;; Load System here Command: (in-package 'ac) #<Package AC 60400002>Command: (autoclass-input-check :data-file "cha:>p>autoclass>sample>imports-85" :header-file "imports-85" :model-file "imports-85") ################# Starting Input Check ################### ############################################################ To log file: CHA:>p>autoclass>sample>IMPORTS-85&IMPORTS-85&IMPORTS-85.LOG During Loading of: [1] CHA:>p>autoclass>sample>imports-85.db2.1, [2] CHA:>p>autoclass>sample>imports-85.hd2.1, and [3] CHA:>p>autoclass>sample>imports-85.model.1. [Attribute #, value # and datum # are zero based.] Loading CHA:>p>autoclass>sample>imports-85.hd2.1 into package AC ADVISORY[2]: No translations were provided for these type = DISCRETE attributes: #0: "symboling" #2: "make" #3: "fuel-type" #4: "aspiration" #5: "num-of-doors" #6: "body-style" #7: "drive-wheels" #8: "engine-location" #14: "engine-type" #15: "num-of-cylinders" #17: "fuel-system" ADVISORY[3]: The DEFAULT model term type, SINGLE-MULTINOMIAL, will be used for these attributes: #2: "make" #3: "fuel-type" #4: "aspiration" #5: "num-of-doors" #6: "body-style" #7: "drive-wheels" #8: "engine-location" #14: "engine-type" #15: "num-of-cylinders" #17: "fuel-system" ################ Input Check Concluded ################### ############################################################ #<ART-Q-1 54255713>#<ART-Q-1 54255715>Command: (multiple-value-setq (search classifications) (autoclass-search :data-file "cha:>p>autoclass>sample>imports-85" :header-file "imports-85" :model-file "imports-85")) ################# Starting Input Check ################### ############################################################ To log file: CHA:>p>autoclass>sample>IMPORTS-85&IMPORTS-85&IMPORTS-85.LOG During Loading of: [1] CHA:>p>autoclass>sample>imports-85.db2.1, [2] CHA:>p>autoclass>sample>imports-85.hd2.1, and [3] CHA:>p>autoclass>sample>imports-85.model.1. [Attribute #, value # and datum # are zero based.] Skipping a reread of CHA:>p>autoclass>sample>imports-85.db2.1 ################ Input Check Concluded ################### ############################################################ WELCOME TO AUTOCLASS. 1) Each time I have finished a new 'trial', or attempt to find a good classification, I will print the number of classes that trial started and ended with, such as 9->7 2) If that trial results in a duplicate of a previous run, I will print 'dup' first. 3) If that trial results in a classification better than any previous, I will print 'best' first. 4) If more than 30 seconds has passed since the last report, and a new classification has been found which is better than any previous ones, I will report on that classification and on the status of the search so far. 5) This report will include an estimate of the time it will take to find another even better classification, and how much better that will be. In addition I will estimate a lower bound on how long it might take to find the very best classification and how much better that might be. 6) If you are warned about too much time in overhead, you may want to change the arguments n-save, min-save-period, min-report-period, log-file-p, search-file-p, results-file-p 7) To quit the search, type a 'q'. 8) The autoclass-search function will return two values: a structure describing the search so far, and a vector of the best 2 classifications. 9) If needed, every 30 minutes I will save the best 2 classifications so far to CHA:>p>autoclass>sample>IMPORTS-85&IMPORTS-85&IMPORTS-85.DUMP and a description of the search to CHA:>p>autoclass>sample>IMPORTS-85&IMPORTS-85&IMPORTS-85.SEARCH 10) A record of this search will be printed to CHA:>p>autoclass>sample>IMPORTS-85&IMPORTS-85&IMPORTS-85.LOG BEGINNING SEARCH at Thr 26 Apr 1990 14:06:33. best1->1 best2->2 3->3 ————— NEW BEST CLASSIFICATION FOUND on try 2 ———————–It has 2 CLASSES with WEIGHTS 108 97 PROBABILITY of both the data and the classification = exp(-17097.9) (Also found 1 other better than last report.) ————— SEARCH STATUS as of Thr 26 Apr 1990 14:07:33 ——————————- It just took 57 seconds to find a classification exp(infinity) times more probable. Estimate < 42 seconds to find a classification exp(226.4) times more probable. Estimate » 57 seconds to find the very best classification, which may be exp(92.3) to exp(17097.9) times more probable. Have seen 3 of the estimated > 6 possible classifications (based on 0 duplicates so far) Log-Normal fit to classifications probabilities has M(ean) -17413.7, S(igma) 421.3 Choosing initial n-classes off of list (5 7 10 15 25) best5->5 ————— NEW BEST CLASSIFICATION FOUND on try 4 ———————–It has 5 CLASSES with WEIGHTS 74 51 33 25 22 PROBABILITY of both the data and the classification = exp(-16649.6) ————— SEARCH STATUS as of Thr 26 Apr 1990 14:08:58 ——————————- It just took 1 minute 29 seconds to find a classification exp(448.3) times more probable. Estimate < 1 minute 55 seconds to find a classification exp(245.1) times more probable. Estimate » 3 minutes 36 seconds to find the very best classification, which may be exp(86.3) = 2.92e+37 to exp(16649.6) times more probable. Have seen 4 of the estimated > 10 possible classifications (based on 0 duplicates so far) Log-Normal fit to classifications probabilities has M(ean) -17222.7, S(igma) 514.1 Choosing initial n-classes off of list (7 10 15 25) 7->7 10->10 15->15 best25->19 ————— NEW BEST CLASSIFICATION FOUND on try 8 ———————–It has 19 CLASSES with WEIGHTS 24 18 17 16 15 14 14 13 11 11 8 8 7 6 6 5 4 4 4 PROBABILITY of both the data and the classification = exp(-16639.9) ————— SEARCH STATUS as of Thr 26 Apr 1990 14:16:02 ——————————- It just took 7 minutes 8 seconds to find a classification exp(9.7) = 1.63e+4 times more probable. Estimate < 8 minutes 24 seconds to find a classification exp(29.5) = 6.30e+12 times more probable. Estimate » 33 minutes 4 seconds to find the very best classification, which may be exp(63.6) = 4.07e+27 to exp(2863.2) times more probable. Have seen 8 of the estimated > 36 possible classifications (based on 0 duplicates so far) Log-Normal fit to classifications probabilities has M(ean) -16723.3, S(igma) 76.8 Choosing initial n-classes randomly from a log-normal [M-S,M,M+S]=[1.9,5.5,15.9] 24->24 5->5 2->2 best6->6 ————— NEW BEST CLASSIFICATION FOUND on try 12 ———————–It has 6 CLASSES with WEIGHTS 69 50 32 24 16 14 PROBABILITY of both the data and the classification = exp(-16620.1) ————— SEARCH STATUS as of Thr 26 Apr 1990 14:21:15 ——————————- It just took 5 minutes 16 seconds to find a classification exp(19.8) = 3.91e+8 times more probable. Estimate < 13 minutes 32 seconds to find a classification exp(78.0) = 7.18e+33 times more probable. Estimate » 1 hour 20 minutes to find the very best classification, which may be exp(250.9) to exp(16620.1) times more probable. Have seen 12 of the estimated > 78 possible classifications (based on 0 duplicates so far) Log-Normal fit to classifications probabilities has M(ean) -16867.4, S(igma) 223.0 Choosing initial n-classes randomly from a log-normal [M-S,M,M+S]=[2.9,7.2,17.7] 13->11 ENDING SEARCH because you asked me to at Thr 26 Apr 1990 14:22:34 after a total of 13 tries over 16 minutes A log of this search is in #P"CHA:>p>autoclass>sample>IMPORTS-85&IMPORTS-85&IMPORTS-85.LOG" Results stored in file #P"CHA:>p>autoclass>sample>IMPORTS-85&IMPORTS-85&IMPORTS-85.DUMP" This search can be restarted using (AUTOCLASS-SEARCH :OUTPUT-FILES-DEFAULT #P"CHA:>p>autoclass>sample>IMPORTS-85&IMPORTS-85&IMPORTS-85") ————— SUMMARY OF 10 BEST RESULTS ————————–PROBABILTY: exp(-16620.1) N-CLASSES: 6 FOUND ON TRY: 12 *SAVED* PROBABILTY: exp(-16639.9) N-CLASSES: 19 FOUND ON TRY: 8 *SAVED* PROBABILTY: exp(-16649.6) N-CLASSES: 5 FOUND ON TRY: 4 PROBABILTY: exp(-16653.6) N-CLASSES: 7 FOUND ON TRY: 5 PROBABILTY: exp(-16770.2) N-CLASSES: 11 FOUND ON TRY: 13 PROBABILTY: exp(-16773.8) N-CLASSES: 15 FOUND ON TRY: 7 PROBABILTY: exp(-16780.2) N-CLASSES: 24 FOUND ON TRY: 9 PROBABILTY: exp(-16895.2) N-CLASSES: 10 FOUND ON TRY: 6 PROBABILTY: exp(-17086.4) N-CLASSES: 2 FOUND ON TRY: 11 PROBABILTY: exp(-17097.9) N-CLASSES: 2 FOUND ON TRY: 2 #S(SEARCH :N 13 :TIME 990 :N-DUPS 0 :N-DUP-TRIES 91 :LAST-TRY-REPORTED-VECTOR #<ART-Q-1 410651530> :TRIES #<ART-Q-13 15000452>) Command: Show Directory (files [default CHA:>p>autoclass>sample>*.*.newest]) CHA:>p>autoclass>sample>*.*.newest CHA:>p>autoclass>sample>*.*.newest 32670 free, 161060/193730 used (83