Databases
StackOverflow2013
Postsdbo
PO
 

Note that there are some explanatory texts on larger screens.

plurals
  1. PO
    primarykey
    data
    text
    <p>You can write fast or slow code in any language :-)</p> <p>Based on a quick inspection of some of the Clojure code, I would say that the main reason for the performance difference is that the Clojure benchmark code hasn't yet been fully optimised to use the fastest language features available.</p> <p>For example the following features in Clojure are all very cool and useful for development convenience, but incur some runtime performance overhead:</p> <ul> <li>Lazy sequences and lists</li> <li>Dynamic Java interoperability using reflection</li> <li>Runtime function composition / first class functions</li> <li>Multimethods / dynamic dispatch</li> <li>Dynamic compilation with eval or on the REPL</li> <li>BigInteger arithmetic</li> </ul> <p>If you want absolute maximum performance (at the cost of some extra complexity), you would want to rewrite code to avoid these and use things like:</p> <ul> <li>Static type hinting (to avoid reflection)</li> <li>Transients</li> <li>Macros (for compile time code manipulation)</li> <li>Protocols</li> <li>Java primitives and arrays</li> <li>loop / recur for iteration</li> </ul> <p>With judicious use of the above, I have found that it is generally possible to get very close to Java performance in Clojure 1.2+, e.g. consider the following code to do one million additions:</p> <p><strong>Unoptimised Clojure</strong> using a lazy sequence and biginterger arithmetic. It's nice and functional but it's not particularly fast:</p> <pre><code>(reduce (fn [acc val] (unchecked-int (unchecked-add (int acc) (int val)))) (range 0 1000000)) =&gt; "Elapsed time: 65.201243 msecs" </code></pre> <p><strong>Optimized Clojure</strong> with primitive arithmetic and loop / recur:</p> <pre><code>(loop [acc (int 0) i (int 0)] (if (&gt;= i (int 1000000)) acc (recur (unchecked-add acc i) (unchecked-inc i)) )) =&gt; "Elapsed time: 0.691474 msecs" </code></pre> <p><strong>Java</strong> code, a pretty standard iterative loop:</p> <pre><code>public static int addMillion() { int result=0; for (int i=0; i&lt;1000000; i++) { result+=i; } return result; } =&gt; "Elapsed time: 0.692081 msecs" </code></pre> <p>p.s. I have used unchecked-add rather than + in the Clojure code so that it matches the integer overflow behaviour of Java.</p>
    singulars
    1. This table or related slice is empty.
    1. POWhy does clojure do worse than scala in the alioth benchmarks
      singulars
      1. PTQuestion
    1. PTAnswer
    1. USmikera
    1. USmikera
    plurals
    1. This table or related slice is empty.
    1. This table or related slice is empty.
    1. POWhy does clojure do worse than scala in the alioth benchmarks
      singulars
      1. PTQuestion
    1. This table or related slice is empty.
    1. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    2. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    3. VO
      singulars
      1. PO
      1. This table or related slice is empty.
      1. VTUpMod
    1. COYou say you've looked at some of the code and then suggest features which "incur some runtime performance overhead" with the implication that those features have something to do with the performance overhead of the programs you looked at - but you don't say which features you think effect which programs. Maybe none of the measured programs use lazy sequences or multimethods or ...
      singulars
      1. PO
      1. USigouy
    2. COYou can demonstrate that it's specifically "possible to get very close to Java performance in Clojure 1.2+" by contributing programs that do ;-)
      singulars
      1. PO
      1. USigouy
    3. COI have a list implementation with fast random access written in Clojure (targeting 1.2) which is exactly the same speed for random access as a vector which is written in Java. So it is possible to get down to the speed of Java. However, it's not a questionable benchmark but something which is useful in real programs. So that maybe doesn't count as demonstration.
      singulars
      1. PO
      1. USkotarak
 

Querying!

 
Guidance
A row detail

Detail views are divided into sections. All the information in the data section comes from columns in the selected row. The other sections display data from other, related rows.

Related data can be related in a to-one or a to-many fashion. Captions of data related in a to-many fashion link to a list view showing a filtered view of the table.

Try moving around until you find a non-empty to-many entry and click on the label to get to one. You can move back to the root by clicking on the database name in the header.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload