User:Lchien1/Final project - Revision history

Lchien1: /* What I Did */

2023-05-06T03:14:03Z

‎What I Did

Lchien1: /* Efficiency of Python Analyser */

2023-05-05T20:19:24Z

‎Efficiency of Python Analyser

Lchien1: /* Efficiency of Prior HFST Analyser */

2023-05-05T20:18:43Z

‎Efficiency of Prior HFST Analyser

Lchien1: /* Results */

2023-05-05T20:17:06Z

‎Results

Lchien1: /* Comparison */

2023-05-05T04:37:38Z

‎Comparison

Lchien1: /* Efficiency of Python Analyser */

2023-05-05T04:32:23Z

‎Efficiency of Python Analyser

Lchien1: /* Results */

2023-05-05T04:31:50Z

‎Results

Lchien1: /* Efficiency of Prior HFST Analyser */

2023-05-05T04:29:29Z

‎Efficiency of Prior HFST Analyser

Lchien1: Undo revision 22155 by Lchien1 (talk)

2023-05-05T04:29:08Z

Undo revision 22155 by Lchien1 (talk)

Lchien1: /* Efficiency of Prior HFST Analyser */

2023-05-05T04:27:56Z

‎Efficiency of Prior HFST Analyser

@@ Line 33: / Line 33: @@
   for each option{
       for each form{
-          append to forms: a copies of the form with options attached
+          append to forms: a copy of the form with options attached
+      }
+  }

@@ Line 61: / Line 61: @@
 **Coverage: 235/2925
 **Time: 1.3 seconds
-**Memory: 63.1 megabytes
+**Memory: 59.4 megabytes
 *For ami.wiki.20230320.txt (2.93 MB)
 **Coverage: 597/50245

@@ Line 69: / Line 69: @@
 ===Efficiency of Prior HFST Analyser===
 To test, use:
-  /usr/bin/time coverage-hfst ../ling073-ami-corpus/ami.corpus.basic.txt ami.automorf.hfst
+  cat ../ling073-ami-corpus/ami.corpus.basic.txt | /usr/bin/time apertium -d . ami-morph
 *For ami.corpus.basic.txt (9.4kb)

@@ Line 46: / Line 46: @@
 The complete output of my analyser can be found in the same github repository, at: https://github.com/lchien1/chien-ami-analysis/blob/main/output/output.txt
-Because my analyser uses for the most part, the same Amis morphemes and patterns that we implemented before using HFST, comparing their coverage is trivial. However, it is still useful to evaluate coverage over a corpus as a sanity check.
+My analyser uses for the most part, the same Amis morphemes and patterns that we implemented before using HFST. However, the coverage analysis for HFST is complex in how it treats individual tokens and what qualifies as a token (numbers? punctuation?). Therefore, comparing their coverage is trivial. However, it is still useful to evaluate coverage over a corpus as a sanity check.
 The Amis corpus can be found at: https://github.swarthmore.edu/Ling073-sp23/ling073-ami-corpus
@@ Line 54: / Line 54: @@
 ===Efficiency of Python Analyser===
 To test efficiency, I used the built in time and resource libraries.
-*For ami.corpus.basic.txt
+*For ami.corpus.basic.txt (9.4kb)
 **Coverage: 62/510 unique forms
 **Time to compile: 0.2 seconds
-**Memory Required: 62.2 megabytes
+**Memory Required: 58.9 megabytes
 ===Efficiency of Prior HFST Analyser===
@@ Line 63: / Line 71: @@
   /usr/bin/time coverage-hfst ../ling073-ami-corpus/ami.corpus.basic.txt ami.automorf.hfst
-*For ami.corpus.basic.txt
+*For ami.corpus.basic.txt (9.4kb)
-**Average time to compile: 0.59 seconds
+**Average time to compile: 0.33 seconds
-**Memory Required: 11.2 kilobytes
+**Memory Required: 12.9 megabytes
 ===Comparison===
-Coverage is not calculated the same way for these two analysers. However, because these two transducers contain mostly the same library of morphemes and similar patterns, it makes sense to compare the efficiency of each against the other. Because the Python program stores the entire dictionary of forms for searching, it is memory inefficient. It turns out, though, that the Python transducer performed better in terms of speed for searching. The time complexity for HFST is O(n) based on the number of letters it has to build. The Python analyzer stores whole forms and searches them, so it is O(n) based on the number of words.
+Coverage is not calculated the same way for these two analysers. However, because these two transducers contain mostly the same library of morphemes and similar patterns, it makes sense to compare the efficiency of each against the other. Because the Python program stores the entire dictionary of forms for searching, it is both memory inefficient and slower. The time complexity for HFST is O(n) based on the number of letters it has to build. The Python analyzer stores whole forms and searches them, so it is O(n) based on the number of words.
 [[Category:sp23_FinalProjects]]

@@ Line 68: / Line 68: @@
 ===Comparison===
-Because these two transducers contain mostly the same library of morphemes and similar patterns, it makes sense to compare the efficiency of each against the other.
+Coverage is not calculated the same way for these two analysers. However, because these two transducers contain mostly the same library of morphemes and similar patterns, it makes sense to compare the efficiency of each against the other. Because the Python program stores the entire dictionary of forms for searching, it is memory inefficient. It turns out, though, that the Python transducer performed better in terms of speed for searching. The time complexity for HFST is O(n) based on the number of letters it has to build. The Python analyzer stores whole forms and searches them, so it is O(n) based on the number of words.
 [[Category:sp23_FinalProjects]]

@@ Line 63: / Line 63: @@
 *For ami.corpus.basic.txt
-**Average time to compile: 0.59 seconds
+**Average time to compile: 0.59 seconds (minimum 0.54 seconds, maximum 0.61 seconds)
-**Memory Required: 11.2 megabytes
+**Memory Required: 11219 kilobytes
 ===Comparison===