I create from a list of strings a list of instances consisting of token feature sequences. Via command line, I can prune those data based on counts, tf-idf etc. (https://github.com/mimno/Mallet/blob/master/src/cc/mallet/classify/tui/Vectors2Vectors.java). But what if I want to do it in Java? How do I have to extend my code?
My target is to remove most common words for LDA topic modeling.
public static InstanceList createInstanceList(List<String> texts) {
ArrayList<Pipe> pipes = new ArrayList<Pipe>();
pipes.add(new CharSequence2TokenSequence());
pipes.add(new TokenSequenceLowercase());
pipes.add(new TokenSequenceRemoveStopwords());
pipes.add(new TokenSequence2FeatureSequence());
InstanceList instanceList = new InstanceList(new SerialPipes(pipes));
instanceList.addThruPipe(new ArrayIterator(texts));
return instanceList;
}
Thank you in advance for your help!
Look at the code that you linked to for examples, starting around line 125. The FeatureCountTool generates term frequency and document frequency information. You can then generate a pruned alphabet and construct a new instance list, as in Vectors2Vectors, or generate a new stoplist Set and reimport the documents from the source files.
Related
I'm currently writing a TSQL (Sybase/Microsoft SQL) to MySQL translator using the ANTLR4 visitor approach.
I'm able to push comments and whitespaces to different channels so that I can use that information later.
What's not super clear is:
how do I get the data back?
and more importantly how do I plug the comments and whitespaces back into my translated MySQL code?
Re: #1, this seems to work to get the list of all tokens including the comments/whitespaces:
public static List<Token> getHiddenTokensFromString(String sqlIn, int hiddenChannel) {
CharStream charStream = CharStreams.fromString(sqlIn);
CaseChangingCharStream upper = new CaseChangingCharStream(charStream, true);
TSqlLexer lexer = new TSqlLexer(upper);
CommonTokenStream commonTokenStream = new CommonTokenStream(lexer, hiddenChannel);
commonTokenStream.fill();
List<Token> hiddenTokens = commonTokenStream.getTokens();
return hiddenTokens;
}
Re #2, what makes it particularly challenging is that as part of the translation, lines of SQL have to be moved around, some lines removed and some lines added.
Any help will be greatly appreciated.
Thanks.
The ANTLR4 lexer creates a number of tokens, each with an index (a running number). Provided you didn't just skip a token, all tokens are available for later inspection, once the parsing step is done, regardless of their channels (the channel is actually just a number property on a token).
So, given you have a token you want to translate, get its index and then ask the token stream for the tokens with the next smaller index or next higher index. These are usually the hidden whitespaces.
Once you have the whitespace token use its start and stop index to get the original text from the char stream. And since you know where you are in the translation process when you do that, it should be easy to know where to insert the original text.
I am trying to export a large feature collection from GEE. I realize that the Python API allows for this more easily than the Java does, but given a time constraint on my research, I'd like to see if I can extract the feature collection in pieces and then append the separate CSV files once exported.
I tried to use a filtering function to perform the task, one that I've seen used before with image collections. Here is a mini example of what I am trying to do
Given a feature collection of 10 spatial points called "points" I tried to create a new feature collection that includes only the first five points:
var points_chunk1 = points.filter(ee.Filter.rangeContains('system:index', 0, 5));
When I execute this function, I receive the following error: "An internal server error has occurred"
I am not sure why this code is not executing as expected. If you know more than I do about this issue, please advise on alternative approaches to splitting my sample, or on where the error in my code lurks.
Many thanks!
system:index is actually ID given by GEE for the feature and it's not supposed to be used like index in an array. I think JS should be enough to export a large featurecollection but there is a way to do what you want to do without relying on system:index as that might not be consistent.
First, it would be a good idea to know the number of features you are dealing with. This is because generally when you use size().getInfo() for large feature collections, the UI can freeze and sometimes the tab becomes unresponsive. Here I have defined chunks and collectionSize. It should be defined in client side as we want to do Export within the loop which is not possible in server size loops. Within the loop, you can simply creating a subset of feature starting from different points by converting the features to list and changing the subset back to feature collection.
var chunk = 1000;
var collectionSize = 10000
for (var i = 0; i<collectionSize;i=i+chunk){
var subset = ee.FeatureCollection(fc.toList(chunk, i));
Export.table.toAsset(subset, "description", "/asset/id")
}
I am using ELKI to cluster data from CSV file
I use
-resulthandler ResultWriter
-out folder/
to save the outputdata
But as an output I have some strange indexes
ID=2138 0.1799 0.2761
ID=2137 0.1797 0.2778
ID=2136 0.1796 0.2787
ID=2109 0.1161 0.2072
ID=2007 0.1139 0.2047
The ID is more than 2000 despite I have less than 100 training samples
DBIDs are internal; the documentation clearly says that you shouldn't make too much assumptions on them because their implementation may change. The only reason they are written to the output at all is because some methods (such as OPTICS) may require cross-referencing objects by this unique ID.
Because they are meant to be unique identifiers, they are usually continuously incremented. The next time you click on "run" in the MiniGUI, you will get the next n IDs... so clearly, you clicked run more than once.
The "Tips & Tricks" in the ELKI DBID documentation probably answer your underlying question - how to use map DBIDs to line numbers of your input file. The best way is to if you want to have object identifiers, assign object identifiers yourself by using an identifier column (and configuring it to be an external identifier).
For further information, see the documentation: https://elki-project.github.io/dev/dbids
I have files in one folder with following naming convention
ClientID_ClientName_Date_Fileextension
12345_Dell_20110103.CSV
I want to extract ClientID from the filename and store that in a variable. I am unsure how I would do this. It seems that a Script Task would suffice but I am do not know how to proceed.
Your options are using Expressions on SSIS Variables or using a Script Task. As a general rule, I prefer Expressions but mentally, I can tell that's a lot of code, or a lot of intertwined variables.
Instead, I'd use the String.Split method in .NET. If you called the Split method for your sample data and provided a delimiter of the underscore _ then you'd receive a 3 element array
12345
Dell
20110103.CSV
Wrap that in a Try Catch block and always grab the second element. Quick and dirty but of course won't address things like 12345_Dell_Quest_20110103.CSV but you didn't ask that question.
Code approximate
string phrase = Dts.Variables["User::CurrentFile"].Value.ToString()
string[] stringSeparators = new string[] {"-"};
string[] words;
try
{
words = phrase.Split(stringSeparators, StringSplitOptions.None);
Dts.Variables["User::ClientName"].Value = words[1];
}
catch
{
; // Do something with this error
}
I was checking out the Named Entity Recognition feature of NLTK. Is it possible to find out which of the extracted keywords is most relevant to the original text? Also, is it possible to know the type (Person / Organization) of the extracted keywords?
If you have a trained tagger, you can first tag your text and then use the NE classifier that comes with NLTK.
The tagged text should be presented as a list
sentence = 'The U.N.'
tagged_sentence = [('The','DT'), ('U.N.', 'NNP')]
Then, the ne classifier would be called like this
nltk.ne_chunk(tagged_sentence)
It returns a Tree. The classified words will appear as Tree nodes inside the main structure.
The result will include if it is a PERSON, ORGANIZATION or GPE.
To find out the most relevant terms, you have to define a measure of "relevance". Usually tf/idf is used but if you are considering only one document, frequency could be enough.
Computing the frequency of each word within a document is easy with NLTK. First you have to load your corpus and once you have load it and have a Text object, simply call:
relevant_terms_sorted_by_freq = nltk.probability.FreqDist(corpus).keys()
Finally, you could filter out all words in relevant_terms_sorted_by_freq that don't belong to a NE list of words.
NLTK offers an online version of a complete book which I find interesting to start with