|
Freethought & Rationalism ArchiveThe archives are read only. |
03-30-2003, 11:41 AM
|
#51 |
|
Veteran Member
Join Date: Mar 2002
Location: anywhere
Posts: 1,976
|
The problem is even worse, since when you take the modulus of FLOOR(RAND()*1000) by lnMax, which is typically not a factor of 1000 (or 1000 and RANDmax), then once again your distribution is biased. There are some numbers that are underepresented. Let me give an example: say I take 0 to 100 modulo 7. There are 15 0's 15 1's and 15 2's whereas all the rest have 14 numbers to them. This is a clear bias.
EDIT: Once again, more than generous with DNAunion's requests. I really have to stop giving so much time to a flawed program and model. 
|
|
|
03-30-2003, 11:56 AM
|
#52 | |
|
Veteran Member
Join Date: Mar 2002
Location: anywhere
Posts: 1,976
|
Quote:
|
|
|
|
|
03-30-2003, 12:18 PM
|
#53 | |
|
Veteran Member
Join Date: Jan 2001
Location: USA
Posts: 1,072
|
Quote:
Let’s look at one example from table 8 for the archaebacteria and see what we find. Looking at the A-B section we see that the amino acid that follows Gly the most often is Glu. So Glu is the most frequently attached amino acid to the C-terminus of Gly. Does the “mirror image” occur? Is Glu most frequently preceded by Gly? No, Gly is not the most frequently attached amino acid to the N-terminus of Glu. Looking across to the B-A entry for Glu we see that Glu, not Gly, is ranked number 1. Is Gly second? No. Is Gly third? No. Gly is fourth. Clearly, in long chains of amino acids, A-B Gly-Glu and B-A Glu-Gly differ somehow. This suggests that in oligopeptides and polypeptides the amino acid linked to the N-terminus of A and/or the one linked to the C-terminus of B can affect the bonding preferences between the two amino acids of concern, A and B. Since an amino acid has a side chain that juts out from its alpha carbon - with some side chains of amino acids being rather bulky (and hence potentially causing steric interference), some being charged (and hence potentially causing electrostatic repulsion or attraction), some being polar, and some being hydrophobic – it seems logical that amino acids bonded to the two of interest could affect the bonding between those two. |
|
|
|
|
03-30-2003, 12:21 PM
|
#54 | |
|
Veteran Member
Join Date: Mar 2002
Location: anywhere
Posts: 1,976
|
Quote:
|
|
|
|
|
03-30-2003, 12:28 PM
|
#55 | |
|
Veteran Member
Join Date: Jan 2001
Location: USA
Posts: 1,072
|
Quote:
Let's see where you stand regarding computer programming. 1) You misread both my programs, incorrectly determining that they calculate EXACTLY X MATCHES when in fact they calculate AT LEAST X MATCHES. 2) You used the term spaghetti code incorrectly. 3) You don't know the difference between C and C++. 4) Your knowledge of the vast many computer programming languages seems to be limited to the long outdated GW BASIC (that must be where you remember the term spaghetti code from, right?). 5) You don't consider Visual FoxPro to be a 'real' programming language (did you know that the Euro-tunnel is run by a VFP program?) |
|
|
|
|
03-30-2003, 12:36 PM
|
#56 | ||
|
Veteran Member
Join Date: Mar 2002
Location: anywhere
Posts: 1,976
|
Quote:
1) it does a test 2) it jumps to the right code (and provides a return address for when this code finishes) Need I remind DNAunion that both C and FoxPro are higher level languages?  Let's take a look at spaghetti code: Code:
nFoundOne = 0;
while (nFoundOne == 0)
{
nIndex = (int) GetRandomNumber(0, nLetteredTiles - 1);
if (nDiscardTilesOnceChosen == 0)
{
// Doesn't matter if the tile has been
// chosen previously because selected
// tiles are placed back into the urn.
nFoundOne = 1;
}
else if (cUrn[nIndex][2] == 'T')
{
// This tile has already been chosen and
// discarded: it can't be selected again.
nFoundOne = 0;
}
else if (cUrn[nIndex][2] == 'F')
{
// This tile has not been chosen previously.
nFoundOne = 1;
}
}
.PS: A definition of spaghetti code: Quote:
|
||
|
|
|
03-30-2003, 12:46 PM
|
#57 | |
|
Regular Member
Join Date: Mar 2002
Location: Nacogdoches, Texas
Posts: 260
|
Quote:
C++ is indeed a superset of C, and requires some additional programming acumen. But it's a bit inane to lord your mastery over the language when you take no advantage whatsoever of the "++" part of it. |
|
|
|
|
03-30-2003, 12:52 PM
|
#58 | |
|
Veteran Member
Join Date: Mar 2002
Location: anywhere
Posts: 1,976
|
Quote:
|
|
|
|
|
03-30-2003, 01:08 PM
|
#59 | ||
|
Veteran Member
Join Date: Jan 2001
Location: USA
Posts: 1,072
|
Quote:
Quote:
What we get is as follows: 0 = 112 occurrences; P(0) = 0.112 1 = 111 occurrences; P(1) = 0.111 2 = 111 occurrences; P(2) = 0.111 3 = 111 occurrences; P(3) = 0.111 4 = 111 occurrences; P(4) = 0.111 5 = 111 occurrences; P(5) = 0.111 6 = 111 occurrences; P(6) = 0.111 7 = 111 occurrences; P(7) = 0.111 8 = 111 occurrences; P(8) = 0.111 Hardly the kind of bias you would have us believe (though you are correct that it does exist, and technically I should eliminate it). Even with that function being called 4,000,000 times (4 calls / iteration * 1,000,000 iterations) that amount of bias won’t amount to anything really meaningful: if things fell perfectly in line with the above probabilities I’d get 4,000 too many 0’s (448,000 instead of 444,000), which out of 4,000,000 calls is only one-tenth of one percent error. Hey, that's a lot better than Rode's being off by at least 11 orders of magnitude! |
||
|
|
|
03-30-2003, 01:17 PM
|
#60 | |||
|
Veteran Member
Join Date: Mar 2002
Location: anywhere
Posts: 1,976
|
Quote:
In any case, you have simultaneously demonstrated in this post: 1) I have indeed shown your code to be flawed. 2) Your code is not in fact generalizable. So, you get a bias in one case when your number of tiles happens to be 9. But what happened to all of your talk about code reusability? Why can't the number of tiles be 7? or 11? or 20? 3) You have no turned the attention once again on just how "accurate" your code is. Remember, I challenged you to show us just how many significant figures we can trust in your code. What shall we believe now? 1 sig fig? 2 sig figs? What is the 95% confidence interval? You reported: Quote:
4) More importantly, you have demonstrated the usefulness of having exact theoretical results with which to compare, especially when you are running probabilistic studies... 5) Which reminds me to ask why you haven't yet verified my answer to your challenge. So is it 5/14? (I bet it is ) Does it seem a little odd that I can do in one line of calculations, what it took you dozens of lines of code to figure out (and still can't decide whether or not it is "accurate")?6) And of course, going back to the original them of my initial reply, I hope it is clear to readers (if not to DNAunion) why DNAunion's critique adds nothing of substance to Rode's paper. First of all, it is quite obvious, as I have shown, that DNAunion does not actually know a) what he is talking about; and b) what Rode is talking about. Secondly, the "new" results are based on touching up a flawed model, which despite four attempts by DNAunion, still fails to account for the double counting of linkages. Third, DNAunion apparently is sticking by his 1e-7 number. He has more or less admitted that this poses no significant challenge to Rode: Quote:
*yawn* Alrighty, I think I've had enough fun for today. |
|||
|
|
| Thread Tools | Search this Thread |
|
