Good information in the [tag:REPL] tag right here on Stack Overflow:

&gt;About read-eval-print-loop
&gt;
&gt;A Read-Eval-Print Loop (REPL) is an interactive interpreter to a programming language. It originated with LISP systems, but many other languages (Python, Ruby, Haskell, Tcl, etc.) use REPL&#39;s to manage interactive sessions. They allow for simple experimentation with a language by bypassing the compile stage of the &quot;code -&gt; compile -&gt; execute&quot; cycle.
&gt;
&gt;There are 4 components to a REPL (named in LISP notation):
&gt;
&gt;  *  A read function, which reads input from the keyboard
&gt;  *  An eval function, which evaluates code passed to it
&gt;  *  A print function, which formats and displays results
&gt;  *  A loop function, which runs the three previous commands until termination


The first Google hit gives you the definition on Wikipedia: REPL stands for [read–eval–print loop](https://en.wikipedia.org/wiki/Read%E2%80%93eval%E2%80%93print_loop):

&gt; A read–eval–print loop (REPL) is a simple, interactive computer programming environment.

In short, it starts an interactive console where you can type in commands and immediately see the result of these commands.

If supported by language, REPL is a interactive way of code or command execution. 

ftp://ftp.cs.utexas.edu/pub/garbage/cs345/schintro-v14/schintro_114.html

I am currently working on a project, a **simple sentiment analyzer** such that there will be **2 and 3 classes** in **separate cases**. I am using a **corpus** that is pretty **rich** in the means of **unique words** (around 200.000). I used **bag-of-words** method for **feature selection** and to reduce the number of **unique features**, an elimination is done due to a **threshold value** of **frequency of occurrence**. The **final set of features** includes around 20.000 features, which is actually a **90% decrease**, but **not enough** for intended **accuracy** of test-prediction. I am using **LibSVM** and **SVM-light** in turn for training and prediction (both **linear** and **RBF kernel**) and also **Python** and **Bash** in general.

The **highest accuracy** observed so far **is around 75%** and I **need at least 90%**. This is the case for **binary classification**. For **multi-class training**, the accuracy falls to **~60%**. I **need at least 90%** at both cases and can not figure how to increase it: via **optimizing training parameters** or **via optimizing feature selection**?

I have read articles about **feature selection** in text classification and what I found is that three different methods are used, which have actually a clear correlation among each other. These methods are as follows:

 - Frequency approach of **bag-of-words** (BOW)
 - **Information Gain** (IG)
 - **X^2 Statistic** (CHI)

The first method is already the one I use, but I use it very simply and need guidance for a better use of it in order to obtain high enough accuracy. I am also lacking knowledge about practical implementations of **IG** and **CHI** and looking for any help to guide me in that way.

Thanks a lot, and if you need any additional info for help, just let me know.

*** 

- @larsmans: **Frequency Threshold**: I am looking for the occurrences of unique words in examples, such that if a word is occurring in different examples frequently enough, it is included in the feature set as a unique feature.   

- @TheManWithNoName: First of all thanks for your effort in explaining the general concerns of document classification. I examined and experimented all the methods you bring forward and others. I found **Proportional Difference** (PD) method the best for feature selection, where features are uni-grams and **Term Presence** (TP) for the weighting (I didn&#39;t understand why you tagged **Term-Frequency-Inverse-Document-Frequency** (TF-IDF) as an indexing method, I rather consider it as a **feature weighting** approach).  **Pre-processing** is also an important aspect for this task as you mentioned. I used certain types of string elimination for refining the data as well as **morphological parsing** and **stemming**. Also note that I am working on **Turkish**, which has **different characteristics** compared to English. Finally, I managed to reach **~88% accuracy** (f-measure) for **binary** classification and **~84%** for **multi-class**. These values are solid proofs of the success of the model I used. This is what I have done so far. Now working on clustering and reduction models, have tried **LDA** and **LSI** and moving on to **moVMF** and maybe **spherical models** (LDA + moVMF), which seems to work better on corpus those have objective nature, like news corpus. If you have any information and guidance on these issues, I will appreciate. I need info especially to setup an interface (python oriented, open-source) between **feature space dimension reduction** methods (LDA, LSI, moVMF etc.) and **clustering methods** (k-means, hierarchical etc.).

Feature Selection and Reduction for Text Classification

I want to edit a `django-rest-framwork` serializer object before it is saved. This is how I currently do it -   

    def upload(request):
        if request.method == &#39;POST&#39;:
            form = ImageForm(request.POST, request.FILES)
            if form.is_valid(): # All validation rules pass
                 obj = form.save(commit=False)
                 obj.user_id = 15
                 obj.save()

How can I do it with a `django-rest-framework` serializer object?

    @api_view([&#39;POST&#39;,&#39;GET&#39;])
    def upload_serializers(request):
        if request.method == &#39;POST&#39;:
             serializer = FilesSerializer(data=request.DATA, files=request.FILES)
             if serializer.is_valid():
                  serializer.save()

Editing django-rest-framework serializer object before save

I saw a reference to creating a &quot;REPL&quot;.  What is a REPL?

    var arDrone = require(&#39;ar-drone&#39;);
    var client  = arDrone.createClient();
    client.createRepl();


What is a &quot;REPL&quot; in javascript?

I saw a reference to creating a "REPL". What is a REPL?
<pre><code class="hljs language-ini">var arDrone = require('ar-drone');
var client = arDrone.createClient();
client.createRepl();
</code></pre>

I am using requireJS 2.x. I found out that some tutorials (and the official docs) sometimes use 

    requirejs.config({ [...] });
    requirejs([&quot;module&quot;]) ...

and sometimes

    require.config({ [...] });
    require([&quot;module&quot;]) ...

Is there any difference between those two functions (require and requirejs)? I could not find any word about that in the docs. :(

RequireJS: Difference between &quot;requirejs&quot; and &quot;require&quot; functions

I know there are a lot of questions about it, but I tried several solutions, and nothing works.

In my django app I have a form:

    &lt;form method=&#39;post&#39;&gt;
        &lt;button type=&#39;submit&#39;&gt;Send&lt;/button&gt;
    &lt;/form&gt;

I wan&#39;t to disable the button once the user has submitted the form. Using other questions, I tried several things, like:
 
    &lt;button type=&#39;submit&#39; onclick=&quot;this.disabled=true&quot;&gt;Send&lt;/button&gt;

When I click, the button is disabled... but the form is not submitted. And for each try I had the same issue: either the button is disabled or the form is submitted. I can&#39;t find how to do both...

I&#39;m using Chrome. Any idea on why I have this problem?  Thank you for your help.

Disable button after submit with jQuery

Is it possible to get the object property name as a string

    person = {};
    person.first_name = &#39;Jack&#39;;
    person.last_name = &#39;Trades&#39;;
    person.address = {};
    person.address.street = &#39;Factory 1&#39;;
    person.address.country = &#39;USA&#39;;

I&#39;d like to use it like this:

    var pn = propName( person.address.country ); // should return &#39;country&#39; or &#39;person.address.country&#39;
    var pn = propName( person.first_name );      // should return &#39;first_name&#39; or &#39;person.first_name&#39;


**NOTE**: this code is exactly what I&#39;m looking for. I understand it sounds even stupid, but it&#39;s not.

This is what I want to do with it.

HTML
    &lt;input type=&quot;text&quot; value=&quot;Jack&quot; id=&quot;id_first_name &quot;/&gt;

    person = {};
    person.id_first_name = &#39;Jack&#39;;
    person.id_last_name = &#39;Trades&#39;;
    person.address = {};
    person.address.id_address = &#39;Factory 1&#39;;
    person.address.id_country = &#39;USA&#39;;


    extPort.postMessage
    (
      {
        message : MSG_ACTION,
        propName( person.first_name ): person.first_name
      }
    };

----------------------**ANSWER**-----------------------

Got it thanks to ibu. He pointed the right way and I used a recursive function

    var res = &#39;&#39;;
    
    function propName(prop, value) {
    	for (var i in prop) {
    		if (typeof prop[i] == &#39;object&#39;) {
    			if (propName(prop[i], value)) {
    				return res;
    			}
    		} else {
    			if (prop[i] == value) {
    				res = i;
    				return res;
    			}
    		}
    	}
    	return undefined;
    }
    
    var pn = propName(person, person.first_name); // returns &#39;first_name&#39;
    var pn = propName(person, person.address.country); // returns &#39;country&#39;

DEMO: http://jsbin.com/iyabal/1/edit

Get object property name as a string

I have a site with a header set to `position: fixed`. On one of my pages, I use `scrollIntoView(true)` on an element. My problem is that when `scrollIntoView` is called, the element gets positioned underneath the header. How would I fix this so that the element is shown just below the header?

I&#39;m using the Bootstrap framework and the header is styled with `navbar navbar-fixed-top`.

Using scrollIntoView with a fixed position header

I am trying to add some text into circle. I have been following example from [a mbostock tutorial][1], but wasn&#39;t able to get the right output.

The code snippet is:

```
var data;
var code;

d3.json(&quot;/json/trace.json&quot;, function(json) {
  data = json;
  console.log(data);
  // get code for visualization
  code = data[&quot;code&quot;];
  alert(code);
  var mainSVG = d3
    .select(&quot;#viz&quot;)
    .append(&quot;svg&quot;)
    .attr(&quot;width&quot;, 900)
    .attr(&quot;height&quot;, 900);
  mainSVG
    .append(&quot;circle&quot;)
    .style(&quot;stroke&quot;, &quot;gray&quot;)
    .style(&quot;fill&quot;, &quot;white&quot;)
    .attr(&quot;r&quot;, 100)
    .attr(&quot;cx&quot;, 300)
    .attr(&quot;cy&quot;, 300);
  circle = mainSVG.selectAll(&quot;circle&quot;).data([code]);
});
```

Any suggestions how to get this work?


  [1]: http://mbostock.github.com/d3/tutorial/circle.html

d3 add text to circle

Well... I&#39;m back to square one. I can&#39;t figure this out for the life of me.

I&#39;m getting the following error:

    FATAL ERROR: JS Allocation failed - process out of memory

I could enumerate the dozens (yes, dozens) of things I&#39;ve tried to get to the root of this problem, but really it would be far too much. So here are the key points:

 - I can only get it to happen on my production server, and my app is large and complicated, so it is proving very difficult to isolate
 - It happens even though heap size &amp; RSS size are both &lt; 200 Mb, which should not be a problem given that the machines (Amazon Cloud, CentOS, m1.large) have 8Gb RAM

**My assumption is that** (because of the 2nd point), a leak is probably not the cause; rather, it seems like there&#39;s probably a SINGLE object that is very large. The following thread backs up this theory:: https://stackoverflow.com/questions/11189940/in-node-js-using-json-stringify-results-in-process-out-of-memory-error 

What I really need is some way to find out what the state of the memory is at the moment the application crashes, or perhaps a stack trace leading up to the FATAL ERROR.

Based upon my assumption above, a 10-minute-old heap dump is insufficient (since the object would have not resided in memory).

Node.js &quot;FATAL ERROR: JS Allocation failed - process out of memory&quot; -- possible to get a stack trace?

I have used CoffeeScript for a while. Now I need to write a npm package, can I write it in CoffeeScript, or I should compile CoffeeScript into JavaScript?

Can I write npm package in CoffeeScript?

I&#39;m trying to write a node.js script that watches for changes in a directory of files, and then prints the files that are changed. How can I modify this script so that it watches a directory (instead of an individual file), and prints the names of the files in the directory as they are changed?

    var fs = require(&#39;fs&#39;),
        sys = require(&#39;sys&#39;);
    var file = &#39;/home/anderson/Desktop/fractal.png&#39;; //this watches a file, but I want to watch a directory instead
    fs.watchFile(file, function(curr, prev) {
        alert(&quot;File was modified.&quot;); //is there some way to print the names of the files in the directory as they are modified?
    });

Watch a folder for changes using node.js, and print file paths when they are changed

I already have read the documentation of Node.js and, unless if I missed something, it does not tell what the parameters contain in certain operations, in particular [`fs.mkdir()`][1]. As you can see in the documentation, it&#39;s not very much.

Currently, I have this code, which tries to create a folder or use an existing one instead:

    fs.mkdir(path,function(e){
        if(!e || (e &amp;&amp; e.code === &#39;EEXIST&#39;)){
            //do something with contents
        } else {
            //debug
            console.log(e);
        }
    });

But I wonder is this the right way to do it? Is checking for the code `EEXIST` the right way to know that the folder already exists? I know I can do `fs.stat()` before making the directory, but that would already be two hits to the filesystem.

Secondly, is there a complete or at least a more detailed documentation of Node.js that contains details as to what error objects contain, what parameters signify etc.

  [1]: http://nodejs.org/api/fs.html#fs_fs_mkdir_path_mode_callback

Node.js create folder or use existing

So I have an application running node js with socket.io as a backend and normal javascript as frontend. My application has a login system which currently simply has the client send its login data as soon as it&#39;s connected.

Now I figured it would be much nicer to have the login data sent along with the handshakeData, so I can directly have the user logged in while connecting (instead of after establishing a connection) respectively refuse authorization when the login data is invalid.

I&#39;m thinking it would be best to put my additional data in the header part of the handshakeData, so any ideas how I could do that? (Without having to modify socket.io if possible, but if it&#39;s the only way I can live with it)



Send custom data along with handshakeData in socket.io?

When debugging Python code at the interactive prompt (REPL), often I&#39;ll write some code which raises an exception, but I haven&#39;t wrapped it in a `try`/`except`, so once the error is raised, I&#39;ve forever lost the exception object.

Often the traceback and error message Python prints out isn&#39;t enough. For example, when fetching a URL, the server might return a 40x error, and you need the content of the response via `error.read()` ... but you haven&#39;t got the error object anymore. For example:

    &gt;&gt;&gt; import urllib2
    &gt;&gt;&gt; f = urllib2.urlopen(&#39;http://example.com/api/?foo=bad-query-string&#39;)
    Traceback (most recent call last):
      File &quot;&lt;stdin&gt;&quot;, line 1, in &lt;module&gt;
      ...
    urllib2.HTTPError: HTTP Error 400: Bad Request

Drat, what did the body of the response say? It probably had valuable error information in it...

I realize it&#39;s usually easy to re-run your code wrapped in a try/except, but that&#39;s not ideal. I also realize that in this specific case if I were using the `requests` library (which doesn&#39;t raise for HTTP errors), I wouldn&#39;t have this problem ... but I&#39;m really wondering if there&#39;s a more general way to get the last exception object at a Python prompt in these cases.


How to get the last exception object after an error is raised at a Python prompt?

I created this function in Python 2.7 with `ipython`:

    def _(v):
        return v

later if I call `_(somevalue)`, I get `_ = somevalue`.

    in[3]: _(3)
    out[3]: 3
    in[4]: print _
    out[4]: 3

The function has disappeared! If I call `_(4)` I get:

```none
TypeError: &#39;int&#39; object is not callable
```

Why? What&#39;s wrong with this function?


Assigning a value to single underscore _ in Python/IPython interpreter

I&#39;ve downloaded **Algebird** and I want to try out few things in the **Scala interpreter** using this library. How do I achieve this? 

How to use third party libraries with Scala REPL?

I read [What&#39;s new in Xcode 6][1]. The article introduces some new feature about Xcode 6, and it says:
&gt;Command Line
&gt;
&gt;Xcode’s debugger includes an interactive version of the Swift language, known as the REPL (Read-Eval-Print-Loop). Use Swift syntax to evaluate and interact with your running app or write new code in a script-like environment. The REPL is available from within LLDB in Xcode’s console, or from Terminal.

I want to know how to get the REPL?


  [1]: https://developer.apple.com/xcode/

How can I use swift in Terminal?

When I type `scala` on the terminal to start the repl, it throws this error

    scala&gt; [init] error: error while loading AnnotatedElement, 
    class file &#39;/usr/lib/jvm/java-8-oracle/jre/lib/rt.jar
    (java/lang/reflect/AnnotatedElement.class)&#39; 
    is broken (bad constant pool tag 15 at byte 2713)

When I hit &lt;kbd&gt;enter&lt;/kbd&gt; and type `println(&quot;hello, world&quot;)`, it again throws this

    error: error while loading CharSequence, 
    class file &#39;/usr/lib/jvm/java-8-oracle/jre/lib/rt.jar
    (java/lang/CharSequence.class)&#39; is broken
    (bad constant pool tag 15 at byte 1501)


I am using `Ubuntu 14.04` and `java -version` gives

    java version &quot;1.8.0_05&quot;
    Java(TM) SE Runtime Environment (build 1.8.0_05-b13)
    Java HotSpot(TM) 64-Bit Server VM (build 25.5-b02, mixed mode)

Content Type	Original Author	Original Content on Stackoverflow
Question	Mark Harrison	View Question on Stackoverflow
Solution 1 - Javascript	tomlogic	View Answer on Stackoverflow
Solution 2 - Javascript	Konrad Rudolph	View Answer on Stackoverflow
Solution 3 - Javascript	raymarch	View Answer on Stackoverflow

What is a "REPL" in javascript?

Javascript Problem Overview

Javascript Solutions

Solution 1 - Javascript

Solution 2 - Javascript

Solution 3 - Javascript

Editing django-rest-framework serializer object before save

Feature Selection and Reduction for Text Classification

Attributions