tgen | Statistical NLG for spoken dialogue systems | Natural Language Processing library

The field in database is allowed to be null, so I fixed it by making the field in the model nullable too.

Here is generalized version

I have something working for at least the tEnum side of things. Since you did not specify your representation of Finite I used my own Finite and Nat.

I have included a full code snippet with an example at the bottom of the post, but will only discuss the generic programming parts, leaving out the reasonably standard construction of Peano arithmetic and various useful theorems about it.

A typeclass is used to keep track of things that can be converted into/out of these finite enums. The important bit here is the default type signatures and the default definitions: these mean that if someone derives EnumFin for a class deriving Generic, they don't have to actually write any code, as these defaults will be used. The defaults use methods from another class, which is implemented for the various kinds of things that GHC.Generics can produce. Notice that both the normal and the default signatures use (n ~ ...) => ... n instead of writing the size of the Finite directly in the type signature; this is because GHC will otherwise detect that the default signatures don't have to match the regular signatures (in the case of a class implementation that defines Size but not fromFin or toFin):

You can convert result to dict like this:

The problem is that you're not using a global variable.

To use a global variable in a function, you have to put the global statement in that function, not at the top level. Because you didn't, the Counter inside doit is a local variable. Any variable that you assign to anywhere in a function is local, unless you have a global (or nonlocal) declaration.

And the first time you use that local Counter is right at the top of the while loop, before you've assigned anything to it. So, it's going to raise an UnboundLocalError immediately.

This exception will be propagated back to the main thread as the result of the future. Which you would have seen, except that you never actually evaluate your futures. You just do this:

The problem is that you iterate through all the lines in your code in the first iteration of your for i in list loop. In the second iteration, e.g. i = 2, the read cursor is still at the end of the file. You have to set it to the first line in each iteration. This can be done with Learn.seek(0):

Is there any parameter to change the amount of output records using terasort?

As far as I understand the source code of TeraSort.java, it seems to implement a custom partitioner, partitioning and sorting the full input. So there is no parameter to change that behavior.

As you have already noticed, the performance difference leads to the comparison: generated array vs JSON.parsed. What we have in both cases: same arrays with same numbers? So the look up performance must be the same? No.

Every Javascript engine has various data type structures for representing same values(numbers, objects, arrays, etc). In most cases optimizer tries to find out the best data type to use. And also often generates some additional meta information, like hidden clases or tags for arrays.

There are several very nice articles about the data types:

• v8-perf/data-types
• v8 performance tips

So why arrays created by JSON.parse are slow? The parser, when creating values, doesn't properly optimize the data structures, and as the result we get untagged arrays with boxed doubles. But we can optimize the arrays afterwards with Array.from and in your case, same as the generated arrays, you get smi arrays with smi numbers. Here is an example based on your sample.