669f31e9d6
TST CountVectorizer with empty vocabulary
2012-10-07 15:32:46 +02:00
749ea70370
ENH let DictVectorizer build a CSR matrix directly and use array.array
...
Prevents copying.
2012-09-29 23:01:27 +02:00
c877742ede
FIX unicode support in count vectorizer. Closes #1098 .
2012-09-20 15:36:38 +01:00
3f8b454b4b
FIX: randomly failing CountVectorizer test
2012-09-01 22:36:43 +03:00
05630b0ae4
ENH added min_df keyword to CountVectorizer, default=2
2012-09-01 14:29:35 +02:00
41564813af
ENH fix merge with char_wb_ngram
2012-08-19 14:40:05 +01:00
bf787ac11f
ENH renamed parameter bounds_n to ngram_range, fixed doctests and tests.
2012-08-19 14:26:09 +01:00
47c156295d
ENH renamed ``min_n`` and ``max_n`` parameters in CountVectorizer to enable gridsearch over them together.
2012-08-19 14:26:09 +01:00
b346184a77
changed 'char_nospace' keyword to shorter and meaningful 'char_wb'
2012-08-14 12:24:58 +02:00
e83fe8bf9a
words for n-grams padded with one space on each side
2012-08-03 00:26:16 +02:00
7db59b1ce8
feature_extraction.text.CountVectorizer analyzer 'char_nospace'
...
creates character n-grams but only inside word boundaries, e.g.
'this is it' to 3-grams: ['thi', 'his', 'is', 'it']
2012-08-02 22:00:41 +02:00
f4882b5c38
DOC + TST vocabulary arg in CountVect docstring
...
Somewhere during the last refactoring, the documentation for the argument
went missing.
Also, check for Mapping subclass instead of hasattr "get" and test with a
few different types.
2012-08-01 15:09:35 +02:00
f6bd86e290
COSMIT cleanup tests with pyflakes
2012-07-21 12:57:49 +02:00
37b440e071
ENH do not fail the test reslying on numpy div 0 warnings if those are not spit out by numpy in general
2012-07-04 00:40:04 -04:00
86cedafa08
ENH make all Estimators default constructible (except SparseCoder)
2012-06-26 14:16:03 +02:00
1ebcc4a486
COSMIT pep8
2012-05-15 21:18:30 +02:00
2727820e6a
P3K: Fixed RuntimeError.message
2012-05-11 11:24:27 +02:00
ad66a25f5a
P3K: Modified RuntimeError message args
2012-05-11 11:21:01 +02:00
3e5f2e5e08
ENH backport "assert_less" and "assert_greater", rename "assert_lower" and use it everywhere :)
2012-05-06 15:43:29 +02:00
e4ce5cf20d
ENH: sort features in dict vectorizer + new doc
2012-04-01 03:13:23 +02:00
20b04689d9
Merge branch 'dictvectorizer'
...
Conflicts:
doc/modules/feature_extraction.rst
2012-03-18 18:09:43 +01:00
7af5f6a549
BUG fix doctests for DictVectorizer (nose 0.X compat)
2012-03-17 14:52:24 +01:00
e05e1a42b9
ENH reduce memory usage of DictVectorizer.transform in sparse case
...
Don't materialize iterable's entire contents.
2012-03-17 12:18:22 +01:00
f3263661f5
TEST more strict test for one-of-K coding in DictVectorizer
2012-03-17 01:19:48 +01:00
c128b147b4
COSMIT rename DictVectorizer source files
...
At the request of @mblondel.
2012-03-16 23:12:41 +01:00
0b15b04424
ENH implement one-of-K/one-hot coding in DictVectorizer
...
inverse_transform and feature_names will return the constructed features,
since the values may have been transformed by a further transformer.
2012-03-16 22:13:42 +01:00
43f1819171
ENH add test for DictVectorizer.restrict
2012-03-09 18:41:43 +01:00
3f5dbe2aa4
ENH test feature_extraction.DictVectorizer
2012-03-08 22:21:53 +01:00
d9fd053033
fixed and inheritance bug in TfidfVectorizer.fit_transform + removed vocabulary backward compat that breaks grid_search
2012-03-07 08:07:47 -08:00
c0541fa03b
Renamed Vectorizer to TfidfVectorizer + deprecation warning
2012-03-07 06:48:56 -08:00
3872d44900
No accent stripping by default + various doc fixes
2012-03-06 02:15:28 +01:00
0d1daad65a
FIX: division by zero errors and negative IDF
2012-03-05 23:25:19 +01:00
1216bc5a6d
Make Vectorizer not inherit from TfidfTransformer while preserving direct gridsearchability
2012-03-05 21:08:03 +01:00
864e4fb155
ENH: remove useless array wrap for feature names + more TF-IDF tests
2012-03-05 20:34:45 +01:00
1f542c29d2
finally the right API with plenty of efficient overrides
2012-03-04 21:58:38 +01:00
267015f36e
switch back to the old vocabulary constructor argument
2012-03-04 20:00:13 +01:00
28471b6d2a
Flatten the combined vectorizer as well
2012-03-04 10:37:16 +01:00
fdb7844bf1
add a test for custom dtype
2012-03-03 19:57:39 +01:00
9f6cdc038c
make CountVectorizer able to output binary occurrence info
2012-03-03 19:53:21 +01:00
38d4332293
factorize feature names array
2012-03-03 19:19:33 +01:00
a510d17a05
Flatten the feature extraction API
2012-03-03 17:20:32 +01:00
aaa6611614
COSMIT pep8
2012-02-17 16:49:48 +01:00
64662c395e
Test now works, testing both the Word and Char analyzers
2012-02-16 07:35:42 +11:00
d961bc2f35
Added a test (that doesn't work yet)
...
unicode_error renamed to decode_error
2012-02-15 21:41:16 +11:00
ad8e4912cb
TST: use assert_true instead of assert + remove some relative imports
2012-02-11 15:43:03 +01:00
654a39cf15
ENH more generic dict-like test in CountVectorizer
...
Includes test.
2012-02-07 17:11:03 +01:00
5e92beb694
ENH sublinear tf scaling in TfidfTransformer
2012-02-04 16:51:40 +01:00
b2239332f5
COSMIT fix ugly import, left over from LinearSVC refactoring
2012-01-17 09:22:58 +01:00
9b0fbab0d1
ENH merge dense/sparse LinearSVC, part 3: deprecate sparse.LinearSVC
2012-01-16 18:27:14 +01:00
ec72437067
FIX: convert lena to float32 (originally it's ints)
...
Fixes exception in recent scipy.
2012-01-09 14:28:22 +01:00
Lars Buitinck