Recursion depth exceeded on evaluating d = d[[1]]

Question

d2 = Import["G:\\mauritius male  death rate(08-10).xlsx"]

{{{0.5, 0.0142}, {2.5, 0.00223}, {7., 0.00158}, {12., 0.00151}, {17., 
   0.0035}, {22., 0.0054}, {27., 0.00751}, {32., 0.01028}, {37., 
   0.01604}, {42., 0.02347}, {47., 0.03576}, {52., 0.05677}, {57., 
   0.07494}, {62., 0.11366}, {67., 0.16382}, {72., 0.23114}, {77., 
   0.32861}, {82., 0.44569}, {85., 0.99999}}}

d = d[[1]]

Recursion depth of 1024 exceeded.<<

Am trying to fit this data in

nlm = NonlinearModelFit[d, b Exp[m x]/(1 + b Exp[m x]), {b, m}, x] // Normal

d = d[[1]] will certainly yield an error. Shouldn't it be d2[[1]]? — J. M.'s missing motivation, Jun 22 '16 at 04:52
when am putting d2=d[[1]] am getting another error Part specification d[[1]] is longer than depth of object. >> — akshay0808, Jun 22 '16 at 05:09
How about this: upload your Excel file somewhere and link to it here, if you want us to be able to help you. — J. M.'s missing motivation, Jun 22 '16 at 05:21
There seems to be two questions here. 1) why OP is getting recursion depth message. 2) How OP can make fit to imported data. Which one is the real question? — m_goldberg, Jun 23 '16 at 01:36

Young · Answer 1 · 2016-06-22T20:42:04.093

4

So, I think this is what you actually wanted:

d2 = {{{0.5, 0.0142}, {2.5, 0.00223}, {7., 0.00158}, {12., 
     0.00151}, {17., 0.0035}, {22., 0.0054}, {27., 0.00751}, {32., 
     0.01028}, {37., 0.01604}, {42., 0.02347}, {47., 0.03576}, {52., 
     0.05677}, {57., 0.07494}, {62., 0.11366}, {67., 0.16382}, {72., 
     0.23114}, {77., 0.32861}, {82., 0.44569}, {85., 0.99999}}};
d = d2[[1]];

logf = LogitModelFit[d, x, x]

Show[Plot[logf[x], {x, 0, 125}, PlotRange -> All], 
 ListPlot[d, PlotRange -> All]]

1/(1+Exp[8.81571 - 0.111638 x]

You can also add weighting to your data.

d1 = {0.1, 0.1, 0.1, 0.1, 0.2, 0.2, 0.2, 0.2, 0.3, 0.3, 
0.3, 0.4, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0};
logfw = LogitModelFit[d, x, x, ConfidenceLevel -> 0.9, Weights -> d1]

Your original equation yields the following (when the initial conditions are set):

nlm = NonlinearModelFit[d, a Exp[b x] /(1 + a Exp[b x]), {{a, 0.001}, {b, 0.001}}, x]

2.88946*10^-7 Exp[0.190243 x] / 1 + 2.88946*10^-7 Exp[0.190243 x]

edited Jun 22 '16 at 20:42

answered Jun 22 '16 at 05:23

Young

7,495
1
20
45

Thank you. But when am doing this its telling me its not a rectangular array? – akshay0808 Jun 22 '16 at 05:26
@akshay0808 Even with the exact same code posted by Young? – MarcoB Jun 22 '16 at 05:30
thanks again mate. It was very helpful. – akshay0808 Jun 22 '16 at 05:34
Its good now. @MarcoB Thanks – akshay0808 Jun 22 '16 at 05:35
thanks for update but am working with logistic regressions – akshay0808 Jun 22 '16 at 07:18
1

I need to erase this post, Can someone help? else i might get plagiarism situation. – akshay0808 Jun 22 '16 at 08:39
4

@akshay, that's not cool. Young here assumed good faith in answering your question, and now you want to have it removed? You should have thought of that first before asking. – J. M.'s missing motivation Jun 22 '16 at 09:29
yes i know but i may land in serious trouble – akshay0808 Jun 22 '16 at 09:35
yes it would be better. Thank you @young – akshay0808 Jun 22 '16 at 14:22
I can't see how this answer would get you into trouble. The only potential for that is in the data you posted. If you want to use this answer, you only need to take care of the site's copyright policy (see below, currently cc by-sa 3.0 with attribution required). – Yves Klett Jun 22 '16 at 16:15
1

And I see that you have included similar data in all of your other posts ... what about those – Young Jun 22 '16 at 16:20
Guys am writing a dissertation on kannisto logistic regression for mauritian mortality and using these information how do i reference the site just in case – akshay0808 Jun 22 '16 at 18:46
1

@akshay0808 http://meta.math.stackexchange.com/questions/8212/citing-stackexchange-postings ... on another note is the answer acceptable? – Young Jun 22 '16 at 20:35
@akshay0808 if you are actually concerned about disclosing something you should not, you can delete the question yourself. I really cant imagine why you would need to cite this in any case. – george2079 Jun 22 '16 at 21:13
@george, actually, since Young's answer has been upvoted, the OP will no longer be able to delete his/her question. – J. M.'s missing motivation Jun 23 '16 at 02:09

Recursion depth exceeded on evaluating d = d[[1]]

1 Answers1