Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.python > #77321 > unrolled thread

I have tried and errored a reasonable amount of times

Back to article view | Back to comp.lang.python

Contents

  I have tried and errored a reasonable amount of times Seymore4Head <Seymore4Head@Hotmail.invalid> - 2014-08-30 14:27 -0400
    Re: I have tried and errored a reasonable amount of times Tim Chase <python.list@tim.thechases.com> - 2014-08-30 13:48 -0500
      Re: I have tried and errored a reasonable amount of times Seymore4Head <Seymore4Head@Hotmail.invalid> - 2014-08-30 14:50 -0400
        Re: I have tried and errored a reasonable amount of times Ned Batchelder <ned@nedbatchelder.com> - 2014-08-30 16:20 -0400
          Re: I have tried and errored a reasonable amount of times Seymore4Head <Seymore4Head@Hotmail.invalid> - 2014-08-30 16:32 -0400
        Re: I have tried and errored a reasonable amount of times Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2014-08-31 18:37 +1000
          Re: I have tried and errored a reasonable amount of times Chris Angelico <rosuav@gmail.com> - 2014-08-31 19:16 +1000
      Re: I have tried and errored a reasonable amount of times Grant Edwards <invalid@invalid.invalid> - 2014-09-02 16:43 +0000
        Re: I have tried and errored a reasonable amount of times Seymore4Head <Seymore4Head@Hotmail.invalid> - 2014-09-02 12:59 -0400
        Re: I have tried and errored a reasonable amount of times Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2014-09-03 11:44 +1000
          Re: I have tried and errored a reasonable amount of times Rustom Mody <rustompmody@gmail.com> - 2014-09-02 20:14 -0700
            Re: I have tried and errored a reasonable amount of times Steven D'Aprano <steve@pearwood.info> - 2014-09-03 07:16 +0000
              Re: I have tried and errored a reasonable amount of times Marko Rauhamaa <marko@pacujo.net> - 2014-09-03 10:44 +0300
                Re: I have tried and errored a reasonable amount of times Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2014-09-04 11:54 +1000
                  Re: I have tried and errored a reasonable amount of times Chris Angelico <rosuav@gmail.com> - 2014-09-04 12:03 +1000
                  Re: I have tried and errored a reasonable amount of times Rustom Mody <rustompmody@gmail.com> - 2014-09-03 19:23 -0700
                  Re: I have tried and errored a reasonable amount of times Marko Rauhamaa <marko@pacujo.net> - 2014-09-04 08:36 +0300
              Re: I have tried and errored a reasonable amount of times Denis McMahon <denismfmcmahon@gmail.com> - 2014-09-04 11:17 +0000
                Re: I have tried and errored a reasonable amount of times Chris Angelico <rosuav@gmail.com> - 2014-09-04 21:42 +1000
                  Re: I have tried and errored a reasonable amount of times Denis McMahon <denismfmcmahon@gmail.com> - 2014-09-05 02:53 +0000
                    Re: I have tried and errored a reasonable amount of times Chris Angelico <rosuav@gmail.com> - 2014-09-05 13:02 +1000
                    Re: I have tried and errored a reasonable amount of times Grant Edwards <invalid@invalid.invalid> - 2014-09-05 14:54 +0000
                      Re: I have tried and errored a reasonable amount of times Marko Rauhamaa <marko@pacujo.net> - 2014-09-05 19:45 +0300
    Re: I have tried and errored a reasonable amount of times Mark Lawrence <breamoreboy@yahoo.co.uk> - 2014-08-30 22:21 +0100
      Re: I have tried and errored a reasonable amount of times Seymore4Head <Seymore4Head@Hotmail.invalid> - 2014-08-30 17:48 -0400
        Re: I have tried and errored a reasonable amount of times Mark Lawrence <breamoreboy@yahoo.co.uk> - 2014-08-30 23:32 +0100
        Re: I have tried and errored a reasonable amount of times Cameron Simpson <cs@zip.com.au> - 2014-08-31 15:30 +1000
        Re: I have tried and errored a reasonable amount of times Steven D'Aprano <steve+comp.lang.python@pearwood.info> - 2014-08-31 19:05 +1000

Page 1 of 2  [1] 2  Next page →

#77321 — I have tried and errored a reasonable amount of times

I really tried to get this without asking for help.

mylist = ["The", "earth", "Revolves", "around", "Sun"]
print (mylist)
for e in mylist:

 # one of these two choices should print something.  Since neither
does, I am missing something subtle.

    if e[0].isupper == False:  
        print ("False")
    if e[0].isupper == True:
        print ("True")    

I am sure in the first , third and fifth choices should be true. Right
now, I am just testing the first letter of each word.

[toc] | [next] | [standalone]

#77323

On 2014-08-30 14:27, Seymore4Head wrote:
> I really tried to get this without asking for help.
> 
> mylist = ["The", "earth", "Revolves", "around", "Sun"]
> print (mylist)
> for e in mylist:
> 
>  # one of these two choices should print something.  Since neither
> does, I am missing something subtle.
> 
>     if e[0].isupper == False:  
>         print ("False")
>     if e[0].isupper == True:
>         print ("True")    
> 
> I am sure in the first , third and fifth choices should be true.
> Right now, I am just testing the first letter of each word.

There's a difference between e[0].isupper which refers to the method
itself, and e[0].isupper() which then calls that method.  Call the
method, and you should be good to go.

-tkc

[toc] | [prev] | [next] | [standalone]

#77324

On Sat, 30 Aug 2014 13:48:09 -0500, Tim Chase
<python.list@tim.thechases.com> wrote:

>On 2014-08-30 14:27, Seymore4Head wrote:
>> I really tried to get this without asking for help.
>> 
>> mylist = ["The", "earth", "Revolves", "around", "Sun"]
>> print (mylist)
>> for e in mylist:
>> 
>>  # one of these two choices should print something.  Since neither
>> does, I am missing something subtle.
>> 
>>     if e[0].isupper == False:  
>>         print ("False")
>>     if e[0].isupper == True:
>>         print ("True")    
>> 
>> I am sure in the first , third and fifth choices should be true.
>> Right now, I am just testing the first letter of each word.
>
>There's a difference between e[0].isupper which refers to the method
>itself, and e[0].isupper() which then calls that method.  Call the
>method, and you should be good to go.
>
>-tkc
>
That works.
Thanks

[toc] | [prev] | [next] | [standalone]

#77331

On 8/30/14 2:50 PM, Seymore4Head wrote:
> On Sat, 30 Aug 2014 13:48:09 -0500, Tim Chase
> <python.list@tim.thechases.com> wrote:
>
>> On 2014-08-30 14:27, Seymore4Head wrote:
>>> I really tried to get this without asking for help.
>>>
>>> mylist = ["The", "earth", "Revolves", "around", "Sun"]
>>> print (mylist)
>>> for e in mylist:
>>>
>>>   # one of these two choices should print something.  Since neither
>>> does, I am missing something subtle.
>>>
>>>      if e[0].isupper == False:
>>>          print ("False")
>>>      if e[0].isupper == True:
>>>          print ("True")
>>>
>>> I am sure in the first , third and fifth choices should be true.
>>> Right now, I am just testing the first letter of each word.
>>
>> There's a difference between e[0].isupper which refers to the method
>> itself, and e[0].isupper() which then calls that method.  Call the
>> method, and you should be good to go.
>>
>> -tkc
>>
> That works.
> Thanks
>

Also, instead of:

     if e[0].isupper() == False:
     if e[0].isupper() == True:

use:

     if not e[0].isupper():
     if e[0].isupper():

It's clearer and reads better.

-- 
Ned Batchelder, http://nedbatchelder.com

[toc] | [prev] | [next] | [standalone]

#77333

On Sat, 30 Aug 2014 16:20:56 -0400, Ned Batchelder
<ned@nedbatchelder.com> wrote:

>On 8/30/14 2:50 PM, Seymore4Head wrote:
>> On Sat, 30 Aug 2014 13:48:09 -0500, Tim Chase
>> <python.list@tim.thechases.com> wrote:
>>
>>> On 2014-08-30 14:27, Seymore4Head wrote:
>>>> I really tried to get this without asking for help.
>>>>
>>>> mylist = ["The", "earth", "Revolves", "around", "Sun"]
>>>> print (mylist)
>>>> for e in mylist:
>>>>
>>>>   # one of these two choices should print something.  Since neither
>>>> does, I am missing something subtle.
>>>>
>>>>      if e[0].isupper == False:
>>>>          print ("False")
>>>>      if e[0].isupper == True:
>>>>          print ("True")
>>>>
>>>> I am sure in the first , third and fifth choices should be true.
>>>> Right now, I am just testing the first letter of each word.
>>>
>>> There's a difference between e[0].isupper which refers to the method
>>> itself, and e[0].isupper() which then calls that method.  Call the
>>> method, and you should be good to go.
>>>
>>> -tkc
>>>
>> That works.
>> Thanks
>>
>
>Also, instead of:
>
>     if e[0].isupper() == False:
>     if e[0].isupper() == True:
>
>use:
>
>     if not e[0].isupper():
>     if e[0].isupper():
>
>It's clearer and reads better.


Thanks

[toc] | [prev] | [next] | [standalone]

#77343

For future reference, here is a hint as to how to debug problems like this,
and a cleaner way to write the code.

Seymore4Head wrote:
> On Sat, 30 Aug 2014 13:48:09 -0500, Tim Chase
>  <python.list@tim.thechases.com> wrote: 
>>>     if e[0].isupper == False:
>>>         print ("False")
>>>     if e[0].isupper == True:
>>>         print ("True")
[...]
>>There's a difference between e[0].isupper which refers to the method
>>itself, and e[0].isupper() which then calls that method.  Call the
>>method, and you should be good to go.
>
> That works.
> Thanks

We've all done it. I think that anyone who tells you they've never forgotten
the brackets on a zero-argument function call is probably lying :-) And the
worst thing is that you can stare and stare and stare at the code for a
week and not see what's wrong.

Here are two suggestions for debugging difficult problems like this:

- Take a break, and look at it with fresh eyes. E.g. work on something 
  else for an hour or three, or go and have lunch.

- Use print() to see the intermediate results:

    a = e[0].isupper
    print(e[0], a, a == False, a == True)

You expect to see something like:

    T True False True

but when you see:

    T <built-in method isupper of str object at 0xb7d44820> False False

at least you will know *what* is going on, even if you're unsure of *why*.



Now, here is a better way to write your code. Since isupper() returns True
or False, there is no need to compare to True or False to get a boolean
value. It's already a boolean value! So we can write:

    if not e[0].isupper():
        print("False")
    if e[0].isupper():
        print("True")


or better still:

    if e[0].isupper():
        print("initial letter is upper case")
    else:
        print("initial letter is lower case")


(More descriptive messages are better than cryptic messages.)


-- 
Steven

[toc] | [prev] | [next] | [standalone]

#77345

On Sun, Aug 31, 2014 at 6:37 PM, Steven D'Aprano
<steve+comp.lang.python@pearwood.info> wrote:
> - Use print() to see the intermediate results:
>
>     a = e[0].isupper
>     print(e[0], a, a == False, a == True)

And I'll add to this: *Copy and paste* the original code to craft this
output statement. I recently was trying to figure out a problem, and
retyped to create the output statements... and I didn't make the same
typo, so it worked fine. For some reason, my eye didn't notice the
error, and I couldn't understand why the intermediate results were
correct and the overall result was wrong. (And no, it wasn't the kind
of 'wrong' that results in a nice tidy exception. It just resulted in
wrong results.)

ChrisA

[toc] | [prev] | [next] | [standalone]

#77442

On 2014-08-30, Tim Chase <python.list@tim.thechases.com> wrote:
> On 2014-08-30 14:27, Seymore4Head wrote:
>> I really tried to get this without asking for help.
>> 
>> mylist = ["The", "earth", "Revolves", "around", "Sun"]
>> print (mylist)
>> for e in mylist:
>> 
>>  # one of these two choices should print something.  Since neither
>> does, I am missing something subtle.
>> 
>>     if e[0].isupper == False:  
>>         print ("False")
>>     if e[0].isupper == True:
>>         print ("True")    
>> 
>> I am sure in the first , third and fifth choices should be true.
>> Right now, I am just testing the first letter of each word.
>
> There's a difference between e[0].isupper which refers to the method
> itself, and e[0].isupper() which then calls that method.  Call the
> method, and you should be good to go.

I missed the beginning of the thread, but Why are you comparing things
to True and False?

What you should do is

     if not e[0].isupper():
        asdf()
        
or
        
     if e[0].isupper():
        qwer()     

-- 
Grant Edwards               grant.b.edwards        Yow! My vaseline is
                                  at               RUNNING...
                              gmail.com            

[toc] | [prev] | [next] | [standalone]

#77443

On Tue, 2 Sep 2014 16:43:09 +0000 (UTC), Grant Edwards
<invalid@invalid.invalid> wrote:

>On 2014-08-30, Tim Chase <python.list@tim.thechases.com> wrote:
>> On 2014-08-30 14:27, Seymore4Head wrote:
>>> I really tried to get this without asking for help.
>>> 
>>> mylist = ["The", "earth", "Revolves", "around", "Sun"]
>>> print (mylist)
>>> for e in mylist:
>>> 
>>>  # one of these two choices should print something.  Since neither
>>> does, I am missing something subtle.
>>> 
>>>     if e[0].isupper == False:  
>>>         print ("False")
>>>     if e[0].isupper == True:
>>>         print ("True")    
>>> 
>>> I am sure in the first , third and fifth choices should be true.
>>> Right now, I am just testing the first letter of each word.
>>
>> There's a difference between e[0].isupper which refers to the method
>> itself, and e[0].isupper() which then calls that method.  Call the
>> method, and you should be good to go.
>
>I missed the beginning of the thread, but Why are you comparing things
>to True and False?
>
>What you should do is
>
>     if not e[0].isupper():
>        asdf()
>        
>or
>        
>     if e[0].isupper():
>        qwer()     

I wasn't really comparing it to True.  I was trying to use it as a
line of code, but since it wasn't working I just tried that to test
it.

Since neither one was working, I knew something was wrong.
BTW I am very new to Python.

Thanks

[toc] | [prev] | [next] | [standalone]

#77460

Grant Edwards wrote:

> I missed the beginning of the thread, but Why are you comparing things
> to True and False?

I don't understand why people do it, but it's *incredibly* common. A couple
of weeks ago at work, I had to (gently, in a friendly manner) mock one of
our most senior and accomplished developers for doing exactly that. He was
suitably chagrined :-)

And then I wondered back to my desk and promptly found myself doing the same
thing :-(

I think it may have something to do with the linguistic idiom in English
where we distinguish between logical statements and logical statements
about logical statements for emphasis. E.g., suppose we have a syllogism:

    Plato is a mortal.
    All mortals must eat.
    Therefore Plato must eat.

and we wanted to analyse it, we might say something like:

    Given the undeniable truth that Plato is, in fact, a mortal, then 
    if the statement `all mortals must eat` is true [i.e. is a true
    statement], then `Plato must eat` likewise is true.

to emphasis that we are making a statement about the truth of the statement
`all mortals must eat`.

I'm not suggesting that this is the only way to talk about the truth or
falsity of statements in English, but it is a very common one:

https://www.google.com.au/search?q=%22if+*+is+true%22


-- 
Steven

[toc] | [prev] | [next] | [standalone]

#77464

On Wednesday, September 3, 2014 7:14:14 AM UTC+5:30, Steven D'Aprano wrote:
> Grant Edwards wrote:

> > I missed the beginning of the thread, but Why are you comparing things
> > to True and False?

> I don't understand why people do it, but it's *incredibly* common. A couple
> of weeks ago at work, I had to (gently, in a friendly manner) mock one of
> our most senior and accomplished developers for doing exactly that. He was
> suitably chagrined :-)

> And then I wondered back to my desk and promptly found myself doing the same
> thing :-(

Logically the boolean domain is a domain over which the natural algebra is
boolean algebra, just like there are arithmetic operators over numbers, string
operators over strings etc.

Intuitively its a different story.  Booleans as an (algebraically
manipulable) domain is just plain weird. Not sure why... Perhaps
because it reifies the structure of our more intimate thinking more
than the other domains?

Also as you point out, language probably has much to do with it.
Dijkstra used to point out 

A ∧ (B ∨ C) ≡ (A ∧ B) ∨ (A ∧ C)
A ∨ (B ∧ C) ≡ (A ∨ B) ∧ (A ∨ C)
look normal enough in this form

Put then into the way engineers do it and they become
A(B + C) = AB + AC
A + BC = (A+B)(A+C)

and Dijkstra would (gleefully!) point out that the engineers were much
more unfamiliar/uneasy with the second than the first.

In effect the perfect symmetry of the boolean operators
get marred in our intuition when we use poor notation.

[toc] | [prev] | [next] | [standalone]

#77468

On Tue, 02 Sep 2014 20:14:51 -0700, Rustom Mody wrote:

> Dijkstra
> used to point out
> 
> A ∧ (B ∨ C) ≡ (A ∧ B) ∨ (A ∧ C) A ∨ (B ∧ C) ≡ (A ∨ B) ∧ (A ∨ C) look
> normal enough in this form
> 
> Put then into the way engineers do it and they become A(B + C) = AB + AC
> A + BC = (A+B)(A+C)

o_O

Who uses + for disjunction (∨ OR) and concatenation for conjunction 
(∧ AND)? That's crazy notation.


-- 
Steven

[toc] | [prev] | [next] | [standalone]

#77472

Steven D'Aprano <steve@pearwood.info>:

> Who uses + for disjunction (∨ OR) and concatenation for conjunction (∧
> AND)? That's crazy notation.

That's the classic Boolean algebraic notation. In basic algebra, the two
interesting operations are "addition" and "multiplication". Boolean math
works like elementary arithmetics, with the exception that

   1 + 1 = 1

I'm guessing the modern symbols ∨ and ∧ were derived from the
set-theoretical analogues, ∪ and ∩, which were also formerly used for
the same purpose.


Marko

[toc] | [prev] | [next] | [standalone]

#77516

Marko Rauhamaa wrote:

> Steven D'Aprano <steve@pearwood.info>:
> 
>> Who uses + for disjunction (∨ OR) and concatenation for conjunction (∧
>> AND)? That's crazy notation.
> 
> That's the classic Boolean algebraic notation. 

Says who? (Apart from you, obviously :-) Since when? I've never seen it in
*any* discussion of Boolean algebra.

Since I answer my own question below (spoiler: George Boole), my questions
are rhetorical.

Mathworld says:

    A Boolean algebra is a mathematical structure that is 
    similar to a Boolean ring, but that is defined using 
    the meet and join operators instead of the usual 
    addition and multiplication operators.

and lists the operators as:

    ^ (logical AND, or "wedge") and v (logical OR, or "vee")

http://mathworld.wolfram.com/BooleanAlgebra.html


Wikipedia lists the operators as 

    And (conjunction), denoted x∧y (sometimes x AND y or Kxy)

    Or (disjunction), denoted x∨y (sometimes x OR y or Axy)

https://en.wikipedia.org/wiki/Boolean_algebra#Operations


So it seems that mathematicians have all but entirely abandoned the old
notation, although they are certainly aware that x∧y is analogous to xy
with both x, y elements of {0, 1}, and similarly for x∨y, although the
analogy is terrible for ∨. 1+1 = 2, not 1. If you perform the addition
modulo 2, then 1+1 = 0, which would make + analogous to XOR, not OR.

I had to go all the way back to George Boole's seminal work "An
Investigation Of The Laws Of Thought" in 1854 to find somebody using that
notation, and he had an excuse, namely he was inventing the subject.

http://gutenberg.org/ebooks/15114

Since Boole isn't with us any more, having died in 1864, my question still
stands: who would be so foolish to use this notation in the 21st century?

Answer: engineers.

http://www.allaboutcircuits.com/vol_4/chpt_7/2.html


-- 
Steven

[toc] | [prev] | [next] | [standalone]

#77519

On Thu, Sep 4, 2014 at 11:54 AM, Steven D'Aprano
<steve+comp.lang.python@pearwood.info> wrote:
> although the
> analogy is terrible for ∨. 1+1 = 2, not 1.

I wouldn't say terrible. Unclear perhaps, but functional. Try this exercise:

false, true = 0, 1 # or use an old Python
if true + true:
    print("true OR true is true")

As long as you accept that any non-zero value is true, and as long as
'and' and 'or' are the only operations you use (you can add 'not' as
long as it's defined the same way: "false if x else true"), addition
for or works fine.

ChrisA

[toc] | [prev] | [next] | [standalone]

#77523

On Thursday, September 4, 2014 7:24:19 AM UTC+5:30, Steven D'Aprano wrote:
> Marko Rauhamaa wrote:

> > Steven D'Aprano:
> >> Who uses + for disjunction (∨ OR) and concatenation for conjunction (∧
> >> AND)? That's crazy notation.
> > That's the classic Boolean algebraic notation. 

> Says who? (Apart from you, obviously :-) Since when? I've never seen it in
> *any* discussion of Boolean algebra.

There are the mathematician/logicians who do boolean algebra (usually
as part of lattice theory). They usually use ∧ ∨ ¬

The guys who (ab)use the add/multiply notations of school algebra are the
electronics books — Karnaugh maps, minimization all that stuff.
One or two pages of truth tables, some of this add/mul notation and then on
its all the standard logic-circuit diagrams for combinational and sequential
circuits.

[toc] | [prev] | [next] | [standalone]

#77536

Steven D'Aprano <steve+comp.lang.python@pearwood.info>:

> Marko Rauhamaa wrote:
>> That's the classic Boolean algebraic notation. 
>
> Says who? (Apart from you, obviously :-) Since when? I've never seen
> it in *any* discussion of Boolean algebra.

I have only run into George Boole, Boolean algebra and booleans in
engineering textbooks (digital electronics, software programming). IIRC,
they *always* used the multiplication and addition symbols.

When studying symbolic logic, I never ran into that notation or Boole or
Boolean algebra or booleans. For example, ∧ and ∨ are called connectors
instead of operations. IOW, they are symbols of a language instead of
functions.


Marko

[toc] | [prev] | [next] | [standalone]

#77542

On Wed, 03 Sep 2014 07:16:34 +0000, Steven D'Aprano wrote:

> Who uses + for disjunction (∨ OR) and concatenation for conjunction (∧
> AND)? That's crazy notation.

The way I was taught it in the mid 1980s, a.b === a and b, a+b === a or b.

-- 
Denis McMahon, denismfmcmahon@gmail.com

[toc] | [prev] | [next] | [standalone]

#77543

On Thu, Sep 4, 2014 at 9:17 PM, Denis McMahon <denismfmcmahon@gmail.com> wrote:
> On Wed, 03 Sep 2014 07:16:34 +0000, Steven D'Aprano wrote:
>
>> Who uses + for disjunction (∨ OR) and concatenation for conjunction (∧
>> AND)? That's crazy notation.
>
> The way I was taught it in the mid 1980s, a.b === a and b, a+b === a or b.

The middle dot is another notation for multiplication, as is abuttal
(not actually concatenation, in this context). So they're all saying
the same thing: boolean 'and' is multiplication, boolean 'or' is
addition.

ChrisA

[toc] | [prev] | [next] | [standalone]

#77571

On Thu, 04 Sep 2014 21:42:56 +1000, Chris Angelico wrote:

> On Thu, Sep 4, 2014 at 9:17 PM, Denis McMahon <denismfmcmahon@gmail.com>
> wrote:
>> On Wed, 03 Sep 2014 07:16:34 +0000, Steven D'Aprano wrote:
>>
>>> Who uses + for disjunction (∨ OR) and concatenation for conjunction (∧
>>> AND)? That's crazy notation.
>>
>> The way I was taught it in the mid 1980s, a.b === a and b, a+b === a or
>> b.
> 
> The middle dot is another notation for multiplication, as is abuttal
> (not actually concatenation, in this context). So they're all saying the
> same thing: boolean 'and' is multiplication, boolean 'or' is addition.

Yes Chris, I know that. I was responding to Stephen's statement that 
using + for or was crazy notation, and pointing out that 30 years ago, 
that's what UK colleges were teaching!

-- 
Denis McMahon, denismfmcmahon@gmail.com

[toc] | [prev] | [next] | [standalone]

Page 1 of 2  [1] 2  Next page →

Back to top | Article view | comp.lang.python

csiph-web