Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: sem
From
William Buchanan <[email protected]>
To
[email protected]
Subject
Re: st: sem
Date
Tue, 26 Feb 2013 07:19:07 -0800
Hi David,
It might not be ideal, but the problem you ran into came from estimating the saturated model. Adding:
satopts(tech(dfp))
Will allow the saturated model to converge (after about 650 iterations). However, the difference in the estimated means of the latent variables is fairly substantial:
first model:
Mean |
Intercept | 26.96308 .5693756 47.36 0.000 25.84713 28.07904
Slope | 10.89299 .2475877 44.00 0.000 10.40773 11.37825
second model:
Mean |
Intercept | 21.83271 .5694494 38.34 0.000 20.71661 22.94881
Slope | 5.13628 .1159911 44.28 0.000 4.908942 5.363619
That being said, it might be an issue that tech support at StataCorp would be in a better position to answer/solve.
HTH,
Billy
On Feb 25, 2013, at 5:36 PM, "Airey, David C" <[email protected]> wrote:
> .
>
> Does anyone find that the recent fix for -sem-:
>
>
> -------- update 25feb2013 -----------------
> 1. sem with method(mlmv) will now attempt to fit the specified model
> even if the saturated model fails to converge.
>
>
> works as advertised? Can someone else confirm on a mac with the
> most recent update it works?
>
> Running my do file below I get the same convergence problem I had
> previously.
>
> clear
>
> input id read1 read2 read3 read4 age1y age2y age3y age4y
> 2 31 47 56 64 7 10 12 14
> 3 36 52 60 75 8 11 13 14
> 13 26 42 53 69 7 9 11 13
> 17 17 37 50 65 6 9 11 13
> 20 27 34 40 47 8 10 12 14
> 21 25 37 41 72 7 9 11 13
> 22 31 49 61 67 8 10 12 14
> 23 32 40 44 59 6 9 11 13
> 26 18 36 46 47 6 8 11 13
> 27 28 28 39 40 8 10 12 14
> 28 21 35 45 51 7 9 11 13
> 29 38 60 60 69 6 9 11 13
> 30 22 40 51 63 7 9 11 13
> 31 72 51 65 76 8 10 12 14
> 32 22 43 60 63 6 9 11 13
> 34 23 49 59 60 7 9 11 13
> 35 19 48 58 74 6 9 11 13
> 36 23 41 57 67 7 9 11 13
> 40 27 41 54 57 8 10 13 14
> 42 18 24 37 54 7 9 12 14
> 44 20 43 50 57 7 9 11 13
> 45 25 43 65 60 7 9 11 13
> 46 16 25 26 37 6 9 11 13
> 47 24 39 44 48 8 10 12 14
> 49 28 50 60 61 6 9 10 12
> 50 36 54 62 77 6 9 11 13
> 51 38 48 54 71 8 10 12 14
> 55 32 36 47 59 8 10 12 14
> 56 26 48 61 74 7 9 11 13
> 57 17 41 54 73 6 8 10 12
> 59 24 37 47 55 8 10 12 14
> 60 23 44 56 72 7 9 11 13
> 62 22 37 48 54 7 9 11 13
> 65 23 45 42 46 7 9 11 14
> 66 23 32 45 53 6 8 11 12
> 67 14 36 52 60 6 9 11 13
> 69 22 33 40 42 7 10 12 14
> 71 18 39 54 58 6 8 11 13
> 72 23 34 45 48 7 9 11 13
> 73 20 29 38 44 7 10 12 13
> 74 19 31 41 51 6 9 11 13
> 76 30 41 50 55 7 9 11 13
> 78 35 42 49 44 8 10 12 14
> 80 18 35 43 54 6 9 11 13
> 81 21 31 39 47 7 10 12 14
> 82 43 53 84 79 7 10 12 14
> 83 21 47 57 52 7 9 11 13
> 84 21 29 45 45 7 9 11 13
> 85 23 38 43 62 6 9 10 12
> 86 18 43 51 72 6 8 11 12
> 87 24 36 46 51 7 10 12 14
> 89 37 50 69 57 8 10 13 14
> 90 20 23 66 58 8 10 12 14
> 91 17 31 54 61 7 9 11 13
> 92 43 56 55 58 7 9 11 13
> 93 18 49 59 70 6 9 11 13
> 95 17 48 58 60 8 10 13 14
> 97 17 37 47 70 6 9 11 13
> 98 47 60 70 80 6 9 11 13
> 101 27 47 59 69 7 9 11 13
> 102 23 42 53 54 7 9 11 13
> 103 18 25 36 47 7 9 11 13
> 106 22 29 34 51 8 10 12 14
> 107 32 36 42 44 7 9 11 13
> 108 22 53 57 61 7 10 12 13
> 109 21 32 49 59 7 9 11 13
> 112 25 40 44 55 6 8 11 13
> 113 18 39 47 58 7 9 11 13
> 115 24 52 58 79 7 10 12 14
> 116 36 38 53 59 6 8 11 12
> 117 16 52 56 66 7 10 12 14
> 118 44 54 58 63 8 10 12 14
> 119 17 40 49 52 7 9 11 13
> 120 40 61 75 66 7 9 11 13
> 121 17 16 22 25 7 9 11 13
> 122 24 50 57 69 7 9 11 13
> 123 21 46 58 68 8 10 12 14
> 124 21 24 34 39 7 10 12 14
> 129 20 40 47 60 7 9 12 14
> 130 18 33 44 61 6 9 11 13
> 134 22 32 46 56 6 9 11 13
> 135 30 47 65 61 7 9 11 13
> 138 24 50 51 58 6 9 11 13
> 139 38 47 63 67 8 10 12 14
> 142 51 57 63 75 8 10 12 14
> 143 32 47 64 74 6 9 11 13
> 144 20 33 40 47 7 9 11 13
> 145 25 39 53 51 7 9 11 13
> 149 18 25 40 41 7 9 11 13
> 150 25 26 37 35 6 9 11 13
> 151 21 34 43 44 7 10 12 14
> 152 17 37 55 63 6 8 10 12
> 153 18 45 56 64 7 9 12 13
> 154 20 23 39 45 6 9 11 13
> 157 35 59 78 71 7 9 11 13
> 158 17 42 48 56 8 10 13 14
> 159 18 33 50 62 8 11 13 14
> 160 17 44 49 58 6 8 11 12
> 161 19 34 35 50 6 9 11 13
> 164 35 50 55 77 7 9 11 13
> 166 18 36 43 51 8 10 12 14
> 171 21 30 37 48 7 9 11 13
> 173 18 25 29 35 7 9 11 13
> 174 21 25 40 33 8 10 13 14
> 175 25 47 53 55 7 9 11 13
> 176 29 45 48 64 7 9 11 13
> 177 19 55 55 55 6 9 11 12
> 178 21 25 31 34 8 10 12 14
> 179 20 36 35 51 7 9 12 13
> 180 28 57 53 58 8 10 12 14
> 182 35 48 59 67 7 10 12 14
> 183 25 51 57 62 6 8 10 12
> 187 18 31 37 43 6 9 11 13
> 188 31 54 66 78 6 8 10 12
> 190 18 24 35 44 6 8 10 12
> 191 22 50 61 81 7 10 12 14
> 195 18 35 45 59 6 8 11 12
> 196 18 39 56 65 8 10 12 14
> 197 23 27 41 44 8 10 13 14
> 199 26 35 41 62 7 10 12 14
> 200 28 38 39 50 7 9 12 13
> 201 30 38 45 55 7 9 11 13
> 202 22 51 60 70 7 10 12 14
> 203 35 43 48 56 8 11 13 14
> 204 18 26 41 40 7 9 11 13
> 207 17 28 39 48 6 9 11 13
> 211 32 50 65 72 7 9 11 13
> 215 16 34 45 57 7 10 12 13
> 217 21 41 49 58 7 10 12 13
> 219 20 36 54 61 6 8 11 12
> 220 35 43 48 59 8 10 12 14
> 221 46 60 61 61 8 10 12 14
> 222 31 52 62 68 7 9 11 13
> 226 19 38 50 48 7 9 11 13
> 227 13 23 25 34 6 9 11 13
> 228 18 32 53 59 8 10 12 14
> 229 23 41 58 69 7 10 12 14
> 231 35 48 58 75 7 10 12 14
> 234 35 48 54 62 8 10 12 14
> 236 32 55 56 49 7 10 12 14
> 240 19 28 33 36 6 9 11 13
> 244 18 35 37 56 6 9 11 13
> 248 31 41 58 74 8 10 12 14
> 250 21 39 50 53 6 9 11 13
> 252 24 36 42 47 7 9 12 13
> 253 19 23 30 28 8 10 12 14
> 255 18 48 65 68 6 9 11 13
> 258 24 38 47 52 7 9 11 13
> 259 33 42 44 66 7 9 11 13
> 262 24 43 55 55 7 10 12 14
> 263 18 38 51 65 6 9 11 13
> 267 33 44 59 58 6 9 11 13
> 269 23 26 31 40 7 9 11 13
> 272 21 41 42 43 8 10 12 14
> 273 15 26 53 56 6 9 11 13
> 275 35 57 61 67 7 9 11 13
> 277 24 54 65 77 7 9 11 13
> 279 12 20 31 34 7 10 11 13
> 280 18 41 53 55 6 9 11 13
> 281 14 22 40 52 6 9 11 13
> 282 27 39 49 55 8 10 12 14
> 287 26 38 58 61 7 9 11 13
> 288 20 39 49 77 6 9 11 13
> 291 37 52 57 78 6 9 11 13
> 295 47 53 61 70 8 10 12 14
> 296 35 57 70 69 8 10 12 14
> 297 26 38 63 61 7 10 12 14
> 299 38 57 62 68 8 10 12 14
> 300 21 30 40 40 7 9 12 13
> 302 18 34 55 64 6 9 11 13
> 303 18 41 50 51 6 8 11 13
> 304 17 32 39 47 7 9 11 13
> 312 20 41 70 80 8 10 12 14
> 313 31 48 48 65 8 10 12 14
> 315 18 49 53 78 7 9 11 13
> 316 26 40 48 51 7 9 11 13
> 318 31 41 47 51 7 10 12 14
> 319 34 46 58 64 8 10 12 14
> 321 28 38 43 50 7 10 12 14
> 322 18 42 51 72 6 8 10 12
> 333 36 56 83 79 7 9 12 13
> 335 19 30 41 44 6 9 11 13
> 337 45 52 61 79 7 9 11 13
> 341 22 47 55 66 7 10 12 14
> 342 22 39 57 60 8 10 12 14
> 344 23 25 29 36 8 10 12 14
> 345 35 40 61 62 6 8 10 12
> 348 26 44 46 50 6 8 11 12
> 350 16 27 40 57 6 9 11 12
> 351 27 49 56 67 7 9 11 13
> 352 18 26 31 40 7 9 11 13
> 353 7 28 37 41 6 8 10 12
> 354 24 32 43 57 8 10 12 14
> 358 18 45 56 58 6 9 11 13
> 359 30 56 51 65 8 10 12 14
> 360 15 38 46 64 6 9 11 13
> 363 35 42 59 63 7 9 11 13
> 368 17 26 31 35 6 9 11 13
> 373 33 49 61 68 8 10 12 14
> 375 22 60 50 56 7 9 12 13
> 378 18 37 44 42 6 9 11 13
> 383 21 55 68 69 7 9 11 13
> 386 40 52 57 80 8 10 12 14
> 388 26 37 46 63 8 10 12 14
> 389 27 40 59 65 7 9 11 13
> 390 20 36 46 55 7 9 11 13
> 392 21 47 70 82 6 9 11 13
> 393 17 21 24 34 8 10 12 14
> 394 28 44 51 60 6 9 11 13
> 395 13 26 34 42 6 9 11 12
> 397 26 25 33 44 6 8 11 12
> 398 56 46 60 63 6 9 11 13
> 399 22 28 34 45 6 9 11 13
> 405 24 39 51 56 7 9 11 13
> end
>
> *** random intercept, random slope, equal error variances
> *** on by wave and by age data; by age has missings as a
> *** result of data restructuring
>
> // works fine
> sem (read1 <- Intercept@1 Slope@0 _cons@0) ///
> (read2 <- Intercept@1 Slope@1 _cons@0) ///
> (read3 <- Intercept@1 Slope@2 _cons@0) ///
> (read4 <- Intercept@1 Slope@3 _cons@0), ///
> latent(Intercept Slope) ///
> cov(Intercept*Slope) ///
> var(e.read1@var e.read2@var e.read3@var e.read4@var) ///
> means(Intercept Slope) ///
> method(ml)
>
> reshape long read@ age@y, i(id) j(wave)
> drop wave
> reshape wide read, i(id) j(agey)
>
> // will not converge
> sem (read6 <- Intercept@1 Slope@0 _cons@0) ///
> (read7 <- Intercept@1 Slope@1 _cons@0) ///
> (read8 <- Intercept@1 Slope@2 _cons@0) ///
> (read9 <- Intercept@1 Slope@3 _cons@0) ///
> (read10 <- Intercept@1 Slope@4 _cons@0) ///
> (read11 <- Intercept@1 Slope@5 _cons@0) ///
> (read12 <- Intercept@1 Slope@6 _cons@0) ///
> (read13 <- Intercept@1 Slope@7 _cons@0) ///
> (read14 <- Intercept@1 Slope@8 _cons@0), ///
> latent(Intercept Slope) ///
> cov(Intercept*Slope) ///
> var(e.read6@var e.read7@var e.read8@var e.read9@var ///
> e.read10@var e.read11@var e.read12@var e.read13@var ///
> e.read14@var) ///
> means(Intercept Slope) ///
> method(mlmv)
>
>
>
>
>
>
>
>
>
>
>
> *
> * For searches and help try:
> * http://www.stata.com/help.cgi?search
> * http://www.stata.com/support/faqs/resources/statalist-faq/
> * http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/help.cgi?search
* http://www.stata.com/support/faqs/resources/statalist-faq/
* http://www.ats.ucla.edu/stat/stata/