[Tutor] How does one construct a module for import?

[Sydney Shall]

Thanks for the detailed reply, Steven.
It seems that you have correctly identified my prblem.

But I am still puzzled, because I do not know how this happened.
I simply copied the two files that I wished to import to a directory 
called (nowMyModule).
It now contains only three files;
 >
pwd
Out[99]: 
u'/Users/sydney/Library/Enthought/Canopy_64bit/User/lib/python2.7/site-packages/MyModule'

ls -l
total 32
-rw-r--r--  1 sydney  staff    29 Jun  2 16:31 __init__.py
-rw-r--r--  1 sydney  staff  2936 Jun  2 16:32 findGraphParametersV2.py
-rw-r--r--  1 sydney  staff  7147 Jun  2 16:32 plotDataV2.py

What does the line above; total 32 refer to?

But when I do the following, this is what I get.


dir(plotDataV2)
Out[107]:
['__builtins__',
  '__doc__',
  '__file__',
  '__name__',
  '__package__',
  'plotData',
  'pylab']


In the file plotDataV2, I have imported of course, pylab.
I do not know where plotData comes from although there is such a file in 
another directory.

dir(findGraphParametersV2)
Out[108]:
['__builtins__',
  '__doc__',
  '__file__',
  '__name__',
  '__package__',
  'findGraphParameters',
  'np',
  'pylab']

Here I have imported numpy as np, pylab and math which doers not appear?

I do understand that I should have only ONE directory containing my own 
Modules and only ONE copy of each file in that directory. But that is 
what I thought I had done.

Do I undestand correctly that what I need to do is to have a single 
directory, say called MyModule, in the directory ....site-packages. And 
then I need to copy just once each of the two function files that I want 
to be importable?

Once again, many thanks for your advice,
Sydney


On 02/06/2014 16:31, Steven D'Aprano wrote:
> On Mon, Jun 02, 2014 at 01:21:29PM +0100, Sydney Shall wrote:
>> I am having a similar problem.
> Actually, I don't think so. Your problem doesn't appear to have anything
> to do with the problem that Charles Agriesti is having. The only
> connection seems to be that you are both using Python. Read on for more
> details.
>
>
>> I have now worked out how to copy my helper file to the correct
>> location, in my case is:
>> '/Users/sydney/Library/Enthought/Canopy_64bit/User/lib/python2.7/site-packages'
>>
>> When I type the following at the IPython prompt I get no error message;
>> import findGraphParametersV2
> Right. This says that, unlike Charles' situation, in your case Python is
> correctly importing your module. You may have a problem, but *importing*
> is not the problem.

>
>
>> And the following led me to believe all was well.
>>
>> I type in the following:
>> In [19]: dir(findGraphParametersV2)
>>
>> Out[19]:
>> ['__builtins__', '__doc__', '__file__', '__name__', '__package__',
>>   'findGraphParameters', 'np', 'pylab']
> This shows that the findGraphParametersV2 (whew, that's a mouthful!)
> module has eight attributes. Some of them (like __name__) are created
> automatically by Python. Others, like np and pylab, are probably created
> when your module imports other modules. The one that you probably care
> about is findGraphParameters, which you need to call using:
>
>      findGraphParametersV2.findGraphParameters( arguments )
>
> Notice that you need to give the module name first, followed by the name
> of the thing inside the module.
>
>   
>> However, when I use the import statement in my program I get a runtime
>> error as follows:
> Just a moment. A couple of sentences ago, you said that importing works.
> Now you say it doesn't. Which is it?
>
> Please be more specific about the code you are running. Unfortunately,
> while we know Python quite well, we're not very good at reading your
> mind, and we can't see your code. You need to identify what line of code
> is being run, and tell us.
>
> If the code is:
>
>      import findGraphParametersV2
>
> which then fails with ImportError, that tells us some valuable
> information. If the code is:
>
>      result = findGraphParametersV2(some, arguments, here)
>
> which then fails with the error you mentioned:
>
>      TypeError: 'module' object is not callable
>
> that tells us something completely different! Both the type of the
> exception (ImportError, TypeError) and the error message are important,
> but equally important is what you did that resulted in the error.
>
>
>> <ipython-input-14-abb1b897e8b9> in <module>()
>> ----> 1 CapitalSimulation(51, 4000.0, 20.0, 20.0, 100, 1.0, 0.0, 40.0, 1.0)
>>
>> /Users/sydney/My_Documents/Political_Economy/Capital_Simulation/Capital/Current
>> version/CapitalWithProdV14.py in CapitalSimulation(number_of_cycles,
>> capital_advanced, unit_constant_capital, wagerate, labour_powers,
>> productivity, prodinc, work_duration, labour_intensity)
>
> If I am reading this correctly, this has absolutely nothing to do with
> the findGraphParametersV2 module or the findGraphParameters function
> inside that module. It looks to me like you have a *different* module,
> called CapitalSimulation, and you try to call it as if it were a
> function.
>
> It is difficult to tell exactly what is going on, but my guess is that
> inside the CapitalSimulation module you have a function *also* called
> CapitalSimulation. So in your module CapitalWithProdV14 (that's the
> THIRD module!!!) you probably have some code like:
>
> import CapitalSimulation  # this is a module, not a function
>
> CapitalSimulation(number_of_cycles, capital_advanced, blah blah blah...)
>
>
> That second line is the problem. You need to change it to:
>
> CapitalSimulation.CapitalSimulation(number_of_cycles, ...)
>
>
> I think. Like I said, without understanding your code, it's difficult to
> be sure exactly what's going on.
>
>
> Reading between the lines, I feel that perhaps somebody has told you
> that you should have one class or one function per file. Or perhaps you
> have been reading Java programming books. Either way, it seems to me
> that you have an excess of modules and too many confusing imports. That
> way leads to frustration.
>
> I believe that you will be much better served to have *one* file per
> project, rather than splitting your project into a dozen itty bitty
> files. That way you don't need to care about importing your own modules,
> because everything is already inside the one file.
>
> If you *must* have separate files, never never never (well, almost
> never) give them the same name as the class or function inside them. A
> good convention is to name the module in all lower case, and the class
> in InitialCaps:
>
> # no, too confusing
> CapitalSimulation.CapitalSimulation(...)
>
> # better
> capital_simulation.CapitalSimulation(...)
>
>
> That way, you can tell at a glance which is the module and which is the
> class inside the module, and if you make a mistake, it will be more
> easily understood:
>
> capital_simulation(...)  # wait a second, that's a module!
>
>
> [...]
>> I do not really understand what Steven is recommending below.
>> Is it an init statement in a file or is it an independent file.
> I don't believe this has anything to do with your problem, but for the
> record, there is no init statement in Python (although classes do have
> an __init__ method). I suggested to Charles that he add an empty
> __init__.py file inside his project folder. That would make it an
> separate file.
>
>
> Good luck!
>
>
>

-- 
Sydney Shall

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/tutor/attachments/20140602/22a44fa4/attachment.html>