| icetindil | 82838a2 | 2013-10-11 16:41:18 -0700 | [diff] [blame] | 1 | #!/bin/bash |
| 2 | # |
| 3 | # Copyright 2010-2011 The Regents of the University of California |
| 4 | # |
| 5 | # Licensed under the Apache License, Version 2.0 (the "License"); you |
| 6 | # may not use this file except in compliance with the License. You |
| 7 | # may obtain a copy of the License at |
| 8 | # |
| 9 | # http://www.apache.org/licenses/LICENSE-2.0 |
| 10 | # |
| 11 | # Unless required by applicable law or agreed to in writing, software |
| 12 | # distributed under the License is distributed on an "AS IS"; BASIS, |
| 13 | # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or |
| 14 | # implied. See the License for the specific language governing |
| 15 | # permissions and limitations under the License. |
| 16 | # |
| 17 | # Author: Rares Vernica <rares (at) ics.uci.edu> |
| 18 | |
| 19 | DATA="/home/rares/data/fuzzyjoin/dblp/records-024" |
| 20 | FUZZYJOIN="/home/rares/fuzzyjoin/fuzzyjoin-core/target/fuzzyjoin-core-0.0.2-SNAPSHOT.jar" |
| 21 | |
| 22 | echo "-- - Step 0: Project and append length - --" |
| 23 | |
| 24 | # java -cp $FUZZYJOIN edu.uci.ics.fuzzyjoin.FuzzyJoinAppendLength $DATA/part-00000 $DATA/part-00000-len |
| 25 | |
| 26 | date |
| 27 | |
| 28 | echo "== START ==" |
| 29 | |
| 30 | echo "-- - Step 1: Sort by length - --" |
| 31 | |
| 32 | # time sort -n -k 5 -t ":" $DATA/part-00000-len > $DATA/part-00000-len-sorted |
| 33 | |
| 34 | echo "-- - Step 2: Tokenize - --" |
| 35 | |
| 36 | # time java -cp $FUZZYJOIN edu.uci.ics.fuzzyjoin.FuzzyJoinTokenize $DATA/part-00000-len-sorted $DATA/part-00000-tokens $DATA/part-00000-tokenized |
| 37 | |
| 38 | echo "-- - Step 3: RID pairs - --" |
| 39 | |
| 40 | time java -Xmx8g -cp $FUZZYJOIN edu.uci.ics.fuzzyjoin.FuzzyJoinMemory .8 $DATA/part-00000-tokenized > $DATA/part-00000-ridpairs |
| 41 | |
| 42 | echo "== END ==" |
| 43 | |
| 44 | date |
| 45 | |
| 46 | |
| 47 | ### SSJoin ### |
| 48 | # cut -d ":" -f 3,4 records-000/part-0000* > ! ssjoin.raw-000/part-00000 |
| 49 | # ~/workspace/ssjoin-bin/txtformat ssjoin.raw-000/part-00000 ssjoin.norm-000/part-00000 l |
| 50 | # sed 's/_\+/ /g' ssjoin.norm-000/part-00000 > ! ssjoin.space-000/part-00000 |
| 51 | # ~/workspace/ssjoin-bin/tokenizer ssjoin.space-000/part-00000 |
| 52 | # ~/workspace/ssjoin-bin/ppjoinplus j .8 ssjoin.space-000/part-00000.bin > /dev/null |
| 53 | # java -jar /fuzzyjoin/fuzzyjoin-core/target/fuzzyjoin-core-0.0.2-SNAPSHOT.jar .8 ssjoin.space-000/part-00000.bin > /dev/null |