Handles the empty string encoding/decoding in Python 2.7 by mstfbl · Pull Request #362 · microsoft/NimbusML

This repository was archived by the owner on Nov 16, 2023. It is now read-only.

.vsts-ci.yml

-Original file line number
+Diff line change
@@ Expand Up / @@ -6,12 +6,6 @@ phases: @@
         name: Windows
         buildScript: build.cmd
         buildMatrix:
-          Py37:
-            _configuration: RlsWinPy3.7
-          Py36:
-            _configuration: RlsWinPy3.6
-          Py35:
-            _configuration: RlsWinPy3.5
           Py27:
             _configuration: RlsWinPy2.7
         buildQueue:
@@ Expand All / @@ -25,6 +19,8 @@ phases: @@
         buildMatrix:
           Py37:
             _configuration: RlsMacPy3.7
+          Py27:
+            _configuration: RlsMacPy2.7
         buildQueue:
           name: Hosted macOS
@@ Expand All / @@ -35,19 +31,6 @@ phases: @@
         name: Linux_Ubuntu16
         buildScript: ./build.sh
         testDistro: ubuntu16
-        buildMatrix:
-          Py37:
-            _configuration: RlsLinPy3.7
-          Py36:
-            _configuration: RlsLinPy3.6
-        buildQueue:
-          name: Hosted Ubuntu 1604
-      # Run tests on CentOS7
-    - template: /build/ci/phase-template.yml
-      parameters:
-        name: Linux_CentOS7
-        buildScript: ./build.sh
-        testDistro: centos7
         buildMatrix:
           Py27:
             _configuration: RlsLinPy2.7
@@ Expand Down @@

src/python/nimbusml/datasets/data/temp_test_data.tsv

-Original file line number
+Diff line change
@@ -0,0 +1,3 @@
+    Sentiment	SentimentText
+	change
+	filler

src/python/nimbusml/datasets/data/train-250.wikipedia.sample.tsv

            
                      Original file line number
                      Diff line number
                      Diff line change
                  
    @@ -189,7 +189,7 @@ Sentiment	SentimentText
  
"  well first, ""accidental suicide"" made me laugh.  There are accidents and you die and then there are suicides and you die.  Second the next sentences hurt my head.  You ASSUME checkers?  I don't.  Some writer is ""theorizing""?  Well this guy believed that George Hodel was the killer of the Black Dahlia.  He has been humiliated for being wrong up and down the internets.  So why not put down MY theory?  Theone in which Martians killed her?  Oh, right, because it is not relevant  "

 ==Cell (film)== Why is it such a horrible thing for me to create a page for the film?  I've seen pages for other movies that are currently in production.  H-E doulbe hocky sticks, I've seen pages for movies that aren't even in production yet.  Can I get some answers, and don't just tell me to read some other WP:BOLOGNA.   

So, in other words, you are professionally on the dole. You must live in your parents basement and leech off of them, like a 11-year old. Maybe if you had a bit of motivation, you could look for a real job, and not play your fantasy as a Wiki boy. I'm sure you couls start a career as a video game player.  

What a joker you are. European parliament has no power to do anything. It is non binding because it is not serious and silly reports like that are not meant to be serious. what is more important is that we ruled your ancestors for centuries and trying to put negative images of turks in the turkey page is not going to change that. This is not a place for you to get your 'revenge'. Go and edit the golden dawn wikipedia because your ideas will only be welcome there.  

your 'revenge'. 

"  == Ban of ""Bryansee"" from Wikipediocracy. ==  Hey, you are Zoloft. The one who banned me from Wikipediocracy with a threat that I die. ""Well"" means dead. ""Recover"" means ""die"". You are wanting me to die by a medication increase or meet my maker. Check this out:    "

MODERATORS ARE SOME OF THE MOST INGORANT AND SELF SERVING JERKS YOU WILL FIND ON THE NET

"  :So I will start with a criticism of the quote from Ollier and Pain, with whom I have more general issues than the ""postorogenic part"". Phrase by phrase that I disagree with: :# Only much later was it realized that the two processes [deformation and the creation of topography] were mostly not closely related, either in origin or in time. Very wrong. Deformation causes topography, and the generation of topography is synchronous with deformation. I will email you a copy of Dahlen and Suppe (1988), which shows that this is the case - send me a message so I have your address and can attach a PDF. They tackle the large-scale deformation of sedimentary rocks via folding and thrusting during orogenesis. :# ...fold-belt mountainous areas...: ""fold-belt"" isn't used professionally (AFAIK) to refer to a collisional mountain-building event. A minor thing though. :# Only in the very youngest, late Cenozoic mountains is there any evident causal relation between rock structure and surface landscape. and the following sentence: If I were British, I would call this ""utter twaddle"". As I mentioned above, there is no way for many of the exposed structures to get to the surface without large amounts of rock uplift and erosion. And as a matter of fact, the trajectory of different units of rock through an orogen is in part determined by patterns of surface erosion. To keep it simple and send you one paper, you'll find this in and at the end of the paper by Dahlen and Suppe (1988). :  "

    @@ -215,7 +215,7 @@ Sentiment	SentimentText
  
I doubt this will get through your thick head (it's not an insult, it's an opinion based on your response) but the problem is not the issue itself. It's that people like you seem to enjoy (whether or not your side gets it right) to discuss, turn, twist and frankly abuse topics like this which are detrimental to the basic goals of Wikis in general and Wikipedia in particular. As John Stewart said to two hacks; You're hurting us. 

2 words learn them SHUT UP DONT FOLLOW ME EVERYWHERE   

" :::hey buddy, hey buddy, guess what? ""I"" dont care realy what ""your"" excuse is, and couldn't care less what Roaringflamer says, but you are obviously obsessed with redirects. If there is anybody that should be banned, its you for vandalism and disruption so there"

"OOOOHHHH With a big long Intellectually Terrifying and Superior name like ""(referenced to Journal of Labelled Compounds and Radiopharmaceuticals)"". How Could the quote be wrong Hey!! How dare I even question it, or possibly be right, in saying the ""supposed"" quote is wrong. What a stupid ignoramus I must be to challenge that.    "

that.    "

 == YOUR THREATENING BEHAVIOUR ==   == YOUR CONSTANT BLOCKING AND SABOTAGE OF MY EDITS IS TANTAMOUNT TO STALIKING. ARE YOU STALKING ME? ARE YOU THREATENING ME STEVE? IS THIS WHAT YOURE ABOUT, THREATENING AND HARRASSING ME? WHY DO YOU KEEP STALKING ME THROUGH WIKIPEDIA? ARE YOU A TWISTED WACKO, DO YOU WISH ME HARM? WHY? WHY ARE YOU HARRASSING ME!!!!!!!!!!!   LEAVE ME ALONE YOU RACIST WACKO!!!!!!!!! ==

 :O: I can't believe you thought that I would call you such a thing. I just wanted to give a cookie so you could get bigger and stronger. Obviously it wasn't because you're a fat pig. I'm sorry for the misunderstanding.   

 It's those biography and political articles you should watch out for.

src/python/nimbusml/datasets/datasets.py

-Original file line number
+Diff line change
@@ Expand Up / @@ -363,6 +363,30 @@ def as_filepath(self): @@
             """
             return os.path.join(DATA_DIR, "train-250.wikipedia.sample.tsv")
+    class TempTestData(DataSet):
+        """
+        TempTestData dataset train.
+        """
+        def __init__(self, inst=None):
+            """
+            Constructor
+            """
+            DataSet.__init__(self, inst=inst)
+            if inst is None:
+                # self.load()
+                pass
+        @property
+        def name(self):
+            return "temp_test_data"
+        def as_filepath(self):
+            """
+            Return file name.
+            """
+            return os.path.join(DATA_DIR, "temp_test_data.tsv")
     class WikiDetox_Test(DataSet):
         """
@@ Expand Down Expand Up / @@ -645,6 +669,7 @@ def as_filepath(self): @@
         topics=lambda: Topics(),
         timeseries=lambda: Timeseries(),
         airquality=lambda: DataSetAirQuality(),
+        temp_test_data=lambda: TempTestData(),
         wiki_detox_train=lambda: WikiDetox_Train(),
         wiki_detox_test=lambda: WikiDetox_Test(),
         gen_twittertrain=lambda: Generated_Twitter_Train(),
@@ Expand Down @@

src/python/nimbusml/tests/ensemble/test_lightgbmclassifier.py

            
                      Original file line number
                      Diff line number
                      Diff line change
                  
    @@ -42,9 +42,9 @@ def test_lightgbmclassifier(self):
  
            accuracy = np.mean(y_test.values.ravel() == scores.values)

            assert_greater(

                accuracy,

                0.58,

                0.55,

                "accuracy should be greater than %s" %

                0.58)

                0.55)

    if __name__ == '__main__':

src/python/nimbusml/tests/feature_extraction/text/test_ngramfeaturizer.py

            
                      Original file line number
                      Diff line number
                      Diff line change
                  
    @@ -1,3 +1,4 @@
  
    # -*- coding: utf-8 -*-

    # --------------------------------------------------------------------------------------------

    # Copyright (c) Microsoft Corporation. All rights reserved.

    # Licensed under the MIT License.

    @@ -7,7 +8,10 @@
  
    import unittest

    import numpy as np

    import sys

    np.set_printoptions(threshold=np.inf)

    import six

    from nimbusml import Pipeline

    from nimbusml.datasets import get_dataset

    from nimbusml.feature_extraction.text import NGramFeaturizer

    from nimbusml.internal.entrypoints._ngramextractor_ngram import n_gram

    @@ -20,24 +24,72 @@ class TestNGramFeaturizer(unittest.TestCase):
  
        def test_ngramfeaturizer(self):

            np.random.seed(0)

            print("hello1")

            train_file = get_dataset('wiki_detox_train').as_filepath()

            print("hello2")

            (train,

             label) = get_X_y(train_file,

                              label_column='Sentiment',

                              sep='\t',

                              encoding="utf-8")

            print("hello3")

            X_train, X_test, y_train, y_test = train_test_split(

                train['SentimentText'], label)

            print("hello4")

            # map text reviews to vector space

            texttransform = NGramFeaturizer(

                word_feature_extractor=n_gram(),

                vector_normalizer='None') << 'SentimentText'

            X_train = texttransform.fit_transform(X_train[:100])

            sum = X_train.iloc[:].sum().sum()

            print("hello5")

            pipe = Pipeline([texttransform])

            print("hello6")

            sentences = X_train[:100].tolist()

            print("hello7")

            #print("X_train Model Just Fit Column Names Start--------------------------------------------------\n")

            pipe.fit(X_train[:100])

            print("hello8")

            print("Name,Size of trained model {},{} bytes".format(pipe.model,os.path.getsize(pipe.model)))

            schema = pipe.get_output_columns()

            print("Schema of pipeline - Len of schema: {}".format(len(schema)))

            #for fea in schema:

            #    print(fea)

            #print("X_train Model Just Fit Column Names End--------------------------------------------------\n")

            #print("X_train Before Just Transform Column Names Start\n")

            X_train_transform = pipe.transform(X_train[:100])

            #print(X_train_transform.iloc[:3])

            #for col in X_train_transform.columns:

            #    print(col)

            #print("X_train Before Just Transform Column Names End\n")

            #print("Len of X_train_transform: {}".format(len(X_train_transform)))

            #X_train = texttransform.fit_transform(X_train[:100])

            #print("X_train Column Names Start\n")

            #for col in X_train.columns:

            #    print(col)

            #print("X_train Column Names End\n")

            #print("X_train_transform.iloc[:].sum() Values Start--------------------------------------------------\n")

            #print(X_train_transform.iloc[:].sum().values)

            #print("X_train_transform.iloc[:].sum() Values End--------------------------------------------------\n")

            print("X_train_transform.iloc[:].sum() IterItems Start--------------------------------------------------\n")

            for col, vals in X_train_transform.iteritems():

                print(col)

                valList = vals.tolist()

                for i in range(len(valList)):

                    if valList[i] > 0:

                        print("Found {} f's for Line {}: {}".format(valList[i], i+1, sentences[i]))

            print("X_train_transform.iloc[:].sum() IterItems End--------------------------------------------------\n")

            sum = X_train_transform.iloc[:].sum().sum()

            print("Sum")

            print(sum)

            assert_equal(sum, 30513, "sum of all features is incorrect!")

            assert_equal(sum, 29565, "sum of all features is incorrect!")

    if __name__ == '__main__':

        unittest.main()

        unittest.main()

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Handles the empty string encoding/decoding in Python 2.7 #362

Uh oh!

Diff view

Diff view

There are no files selected for viewing