galvanizedatascience / george Goto Github PK

cd nlp-feature-selection
$ george check-sprint
# ...
======================================================================
FAIL: runTest (george._helpers.FooTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/tlevine/Zipfian/george/george/_helpers.py", line 7, in runTest
    func()
  File "/home/tlevine/Zipfian/george/george/check_sprint.py", line 31, in <lambda>
    add_test(_suite, lambda: check_readme_has_section(sectionname))
  File "/home/tlevine/Zipfian/george/george/check_sprint.py", line 79, in check_readme_has_section
    n.assert_in('\n## %s' % section, readme)
AssertionError: '\n## Glossary' not found in "This is an introduction to NLP that doesn't discuss machine learning.\n\nOutline\n----------\nI think we should break this into two or three sprints or do it after another\nsprint that has already touched on some of these concepts.\n\n1. Motivate the applications of NLP\n  * Sentiment analysis\n  * Autocorrect\n  * Information retrieval and question-answering\n  * Machine translation\n  * SPAM\n  * ...\n2. Fundamental paradigms\n  * Bags of words\n  * N-grams\n  * String similarity metrics\n  * Tokenization\n3. Examples of the use of the above paradigms in practice\n\nExercise\n----------\nHere's a shorter one that naturally keeps a tabular structure:\nWe have a dataset for which each row has an associated text file.\nStudents extract some simple features from each and write about\ntheir methodology. I'm thinking things like\n* Number of sentences and paragraphs\n* Prominent named entities\n* Prominent N-grams (Yelp does this.)\n* An interesting paragraph or section. For example,\n  * Given some emails, find the salutation and the valediction.\n  * Given some planning licence applications, find the paragraph that\n      mentions the size of the proposed development.\n  * Given some documents of highly variable structure, find the\n      postal addresses\n\nAnother possibility: Students are given a bunch of documents to group.\nHere are some approaches\n* String similarity\n* Topic modelling\n* Extract any interesting features, then use any unsupervised\n    classification algorithm.\n\n"

======================================================================
FAIL: runTest (george._helpers.FooTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/tlevine/Zipfian/george/george/_helpers.py", line 7, in runTest
    func()
  File "/home/tlevine/Zipfian/george/george/check_sprint.py", line 31, in <lambda>
    add_test(_suite, lambda: check_readme_has_section(sectionname))
  File "/home/tlevine/Zipfian/george/george/check_sprint.py", line 79, in check_readme_has_section
    n.assert_in('\n## %s' % section, readme)
AssertionError: '\n## Glossary' not found in "This is an introduction to NLP that doesn't discuss machine learning.\n\nOutline\n----------\nI think we should break this into two or three sprints or do it after another\nsprint that has already touched on some of these concepts.\n\n1. Motivate the applications of NLP\n  * Sentiment analysis\n  * Autocorrect\n  * Information retrieval and question-answering\n  * Machine translation\n  * SPAM\n  * ...\n2. Fundamental paradigms\n  * Bags of words\n  * N-grams\n  * String similarity metrics\n  * Tokenization\n3. Examples of the use of the above paradigms in practice\n\nExercise\n----------\nHere's a shorter one that naturally keeps a tabular structure:\nWe have a dataset for which each row has an associated text file.\nStudents extract some simple features from each and write about\ntheir methodology. I'm thinking things like\n* Number of sentences and paragraphs\n* Prominent named entities\n* Prominent N-grams (Yelp does this.)\n* An interesting paragraph or section. For example,\n  * Given some emails, find the salutation and the valediction.\n  * Given some planning licence applications, find the paragraph that\n      mentions the size of the proposed development.\n  * Given some documents of highly variable structure, find the\n      postal addresses\n\nAnother possibility: Students are given a bunch of documents to group.\nHere are some approaches\n* String similarity\n* Topic modelling\n* Extract any interesting features, then use any unsupervised\n    classification algorithm.\n\n"

======================================================================
FAIL: runTest (george._helpers.FooTestCase)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/tlevine/Zipfian/george/george/_helpers.py", line 7, in runTest
    func()
  File "/home/tlevine/Zipfian/george/george/check_sprint.py", line 31, in <lambda>
    add_test(_suite, lambda: check_readme_has_section(sectionname))
  File "/home/tlevine/Zipfian/george/george/check_sprint.py", line 79, in check_readme_has_section
    n.assert_in('\n## %s' % section, readme)
AssertionError: '\n## Glossary' not found in "This is an introduction to NLP that doesn't discuss machine learning.\n\nOutline\n----------\nI think we should break this into two or three sprints or do it after another\nsprint that has already touched on some of these concepts.\n\n1. Motivate the applications of NLP\n  * Sentiment analysis\n  * Autocorrect\n  * Information retrieval and question-answering\n  * Machine translation\n  * SPAM\n  * ...\n2. Fundamental paradigms\n  * Bags of words\n  * N-grams\n  * String similarity metrics\n  * Tokenization\n3. Examples of the use of the above paradigms in practice\n\nExercise\n----------\nHere's a shorter one that naturally keeps a tabular structure:\nWe have a dataset for which each row has an associated text file.\nStudents extract some simple features from each and write about\ntheir methodology. I'm thinking things like\n* Number of sentences and paragraphs\n* Prominent named entities\n* Prominent N-grams (Yelp does this.)\n* An interesting paragraph or section. For example,\n  * Given some emails, find the salutation and the valediction.\n  * Given some planning licence applications, find the paragraph that\n      mentions the size of the proposed development.\n  * Given some documents of highly variable structure, find the\n      postal addresses\n\nAnother possibility: Students are given a bunch of documents to group.\nHere are some approaches\n* String similarity\n* Topic modelling\n* Extract any interesting features, then use any unsupervised\n    classification algorithm.\n\n"

----------------------------------------------------------------------
Ran 15 tests in 0.013s

FAILED (failures=6)