How to debug this problem? And why it has different bugs for Mac OS and Windows OS?

330 views

Skip to first unread message

Jiaqi Zhang

unread,

Jan 15, 2016, 1:25:10 PM1/15/16

to Stan users mailing list

Hi, All,

I am pretty new to Stan language although I used to have some experience in WinBUGs, but I am still learning the Stan language in R, with RStan.

Here is my detailed problem:

I am trying to examine guessing as a student characteristic in Rasch model. Rather than assuming guessing is an item parameter and is a constant across all student, I assume, more realistic, that guessing varies based on students' latent ability --- the higher the ability, the less chance for guessing. Then whether student answer one particular item correctly, is the weighted combination of guessing and their ability:

Bernoulli (gurssing_j*1/k + (1-guessing_j)*(exp(theta_j - item.diff_i)/(1+exp(theta_j - item.diff_i)))

where i is the number of items, and j is the number of students.

I generated the data in the way that I proposed [In the attached R file, line 23 - 75 is the function that I generated the data], and want to analysis the data. I think this model cannot be analyzed in traditional IRT models, because now guessing is now at the student side. Then I am trying to analysis the data with Rstan using the Bayesian approach.

So I write my model in Stan language, and try to compile it. At a start point, I started with simplest prior distribution for all parameters, and the model:

for(i in 1:n_student)

{

for (j in 1:n_item)

{

y[i,j] ~ bernoulli(guessing[i]*.2+(1-guessing[i])*inv_logit(theta[i] - item_diff[j]));

}

The Rstan model is actually compiled, but there is no samples when I am calling it. And I tried the same code in R on both Mac OS and Windows OS, and it turns out two different problem, but both of them has no sampling in the chain.

For R in Windows OS:

The model is compiled and actually run, no error message, but the model contains no samples. The message shows in the console is (with sessionInfo):

> fit_3pl <- stan(model_code = code_mm, data = init, iter = 2000, chains = 1)

SAMPLING FOR MODEL 'a9de5b0d567a006cf80adb5bae39f3d8' NOW (CHAIN 1).

[1] "Rejecting initial value:"

[2] " Error evaluating the log probability at the initial value."

[3] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[4] "Exception thrown at line 25: stan::math::bernoulli_log: Probability parameter is -0.406224, but must be between (0, 1)"

[5] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[6] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[7] "Rejecting initial value:"

[8] " Error evaluating the log probability at the initial value."

[9] "Initialization between (-2, 2) failed after 100 attempts. "

[10] " Try specifying initial values, reducing ranges of constrained values, or reparameterizing the model."

[1] "error occurred during calling the sampler; sampling not done"

> proc.time() - star

user system elapsed

1.23 0.06 57.49

> fit_3pl

Stan model 'a9de5b0d567a006cf80adb5bae39f3d8' does not contain samples.

For more information, please see the attached IRT_Windows_OS.R file.

> sessionInfo()

R version 3.2.2 (2015-08-14)

Platform: x86_64-w64-mingw32/x64 (64-bit)

Running under: Windows 7 x64 (build 7601) Service Pack 1

locale:

[1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252 LC_MONETARY=English_United States.1252 LC_NUMERIC=C

[5] LC_TIME=English_United States.1252

attached base packages:

[1] stats graphics grDevices utils datasets methods base

other attached packages:

[1] rstan_2.8.0 ggplot2_1.0.1 Rcpp_0.12.1 mvtnorm_1.0-3

loaded via a namespace (and not attached):

[1] codetools_0.2-14 digest_0.6.8 MASS_7.3-44 grid_3.2.2 plyr_1.8.3 gtable_0.1.2 stats4_3.2.2 magrittr_1.5 scales_0.3.0 stringi_1.0-1

[11] reshape2_1.4.1 proto_0.3-10 tools_3.2.2 stringr_1.0.0 munsell_0.4.2 inline_0.3.14 colorspace_1.2-6 gridExtra_2.0.0

For R in Mac OS:

The console returns a more crazy message with only one line of error message (with sessionInfo), for more information, please see the attached IRT_Mac_OS.R file:

> star <- proc.time()

> #set_cppo(mode = "debug")

> fit_3pl <- stan(model_code = code_mm, data = init, iter = 2000, chains = 1)

clang: warning: optimization flag '-ffat-lto-objects' is not supported

clang: warning: argument unused during compilation: '-ffat-lto-objects'

SAMPLING FOR MODEL '3d8fca4d9b3c657abf5a65e146670733' NOW (CHAIN 1).

[1] "Error : "

[2] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[3] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.144568, but must be between (0, 1)"

[4] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[5] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[6] "Rejecting initial value:"

[7] " Error evaluating the log probability at the initial value."

[8] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[9] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.87755, but must be between (0, 1)"

[10] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[11] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[12] "Rejecting initial value:"

[13] " Error evaluating the log probability at the initial value."

[14] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[15] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0211564, but must be between (0, 1)"

[16] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[17] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[18] "Rejecting initial value:"

[19] " Error evaluating the log probability at the initial value."

[20] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[21] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.07851, but must be between (0, 1)"

[22] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[23] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[24] "Rejecting initial value:"

[25] " Error evaluating the log probability at the initial value."

[26] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[27] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.78473, but must be between (0, 1)"

[28] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[29] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[30] "Rejecting initial value:"

[31] " Error evaluating the log probability at the initial value."

[32] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[33] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.165628, but must be between (0, 1)"

[34] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[35] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[36] "Rejecting initial value:"

[37] " Error evaluating the log probability at the initial value."

[38] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[39] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.09169, but must be between (0, 1)"

[40] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[41] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[42] "Rejecting initial value:"

[43] " Error evaluating the log probability at the initial value."

[44] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[45] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0460136, but must be between (0, 1)"

[46] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[47] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[48] "Rejecting initial value:"

[49] " Error evaluating the log probability at the initial value."

[50] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[51] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.23847, but must be between (0, 1)"

[52] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[53] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[54] "Rejecting initial value:"

[55] " Error evaluating the log probability at the initial value."

[56] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[57] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.304645, but must be between (0, 1)"

[58] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[59] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[60] "Rejecting initial value:"

[61] " Error evaluating the log probability at the initial value."

[62] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[63] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.09657, but must be between (0, 1)"

[64] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[65] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[66] "Rejecting initial value:"

[67] " Error evaluating the log probability at the initial value."

[68] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[69] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.84307, but must be between (0, 1)"

[70] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[71] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[72] "Rejecting initial value:"

[73] " Error evaluating the log probability at the initial value."

[74] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[75] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.23891, but must be between (0, 1)"

[76] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[77] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[78] "Rejecting initial value:"

[79] " Error evaluating the log probability at the initial value."

[80] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[81] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.04192, but must be between (0, 1)"

[82] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[83] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[84] "Rejecting initial value:"

[85] " Error evaluating the log probability at the initial value."

[86] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[87] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0759067, but must be between (0, 1)"

[88] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[89] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[90] "Rejecting initial value:"

[91] " Error evaluating the log probability at the initial value."

[92] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[93] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.51935, but must be between (0, 1)"

[94] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[95] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[96] "Rejecting initial value:"

[97] " Error evaluating the log probability at the initial value."

[98] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[99] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.203562, but must be between (0, 1)"

[100] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[101] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[102] "Rejecting initial value:"

[103] " Error evaluating the log probability at the initial value."

[104] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[105] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0408336, but must be between (0, 1)"

[106] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[107] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[108] "Rejecting initial value:"

[109] " Error evaluating the log probability at the initial value."

[110] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[111] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.261902, but must be between (0, 1)"

[112] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[113] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[114] "Rejecting initial value:"

[115] " Error evaluating the log probability at the initial value."

[116] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[117] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0550278, but must be between (0, 1)"

[118] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[119] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[120] "Rejecting initial value:"

[121] " Error evaluating the log probability at the initial value."

[122] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[123] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.041055, but must be between (0, 1)"

[124] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[125] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[126] "Rejecting initial value:"

[127] " Error evaluating the log probability at the initial value."

[128] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[129] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.11561, but must be between (0, 1)"

[130] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[131] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[132] "Rejecting initial value:"

[133] " Error evaluating the log probability at the initial value."

[134] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[135] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.25464, but must be between (0, 1)"

[136] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[137] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[138] "Rejecting initial value:"

[139] " Error evaluating the log probability at the initial value."

[140] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[141] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0928572, but must be between (0, 1)"

[142] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[143] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[144] "Rejecting initial value:"

[145] " Error evaluating the log probability at the initial value."

[146] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[147] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.92796, but must be between (0, 1)"

[148] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[149] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[150] "Rejecting initial value:"

[151] " Error evaluating the log probability at the initial value."

[152] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[153] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.18947, but must be between (0, 1)"

[154] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[155] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[156] "Rejecting initial value:"

[157] " Error evaluating the log probability at the initial value."

[158] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[159] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.155538, but must be between (0, 1)"

[160] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[161] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[162] "Rejecting initial value:"

[163] " Error evaluating the log probability at the initial value."

[164] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[165] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0565987, but must be between (0, 1)"

[166] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[167] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[168] "Rejecting initial value:"

[169] " Error evaluating the log probability at the initial value."

[170] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[171] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.01561, but must be between (0, 1)"

[172] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[173] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[174] "Rejecting initial value:"

[175] " Error evaluating the log probability at the initial value."

[176] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[177] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 2.31113, but must be between (0, 1)"

[178] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[179] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[180] "Rejecting initial value:"

[181] " Error evaluating the log probability at the initial value."

[182] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[183] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.48071, but must be between (0, 1)"

[184] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[185] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[186] "Rejecting initial value:"

[187] " Error evaluating the log probability at the initial value."

[188] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[189] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0183245, but must be between (0, 1)"

[190] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[191] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[192] "Rejecting initial value:"

[193] " Error evaluating the log probability at the initial value."

[194] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[195] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0505354, but must be between (0, 1)"

[196] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[197] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[198] "Rejecting initial value:"

[199] " Error evaluating the log probability at the initial value."

[200] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[201] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.10699, but must be between (0, 1)"

[202] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[203] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[204] "Rejecting initial value:"

[205] " Error evaluating the log probability at the initial value."

[206] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[207] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0787345, but must be between (0, 1)"

[208] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[209] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[210] "Rejecting initial value:"

[211] " Error evaluating the log probability at the initial value."

[212] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[213] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.58232, but must be between (0, 1)"

[214] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[215] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[216] "Rejecting initial value:"

[217] " Error evaluating the log probability at the initial value."

[218] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[219] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.01703, but must be between (0, 1)"

[220] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[221] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[222] "Rejecting initial value:"

[223] " Error evaluating the log probability at the initial value."

[224] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[225] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0209495, but must be between (0, 1)"

[226] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[227] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[228] "Rejecting initial value:"

[229] " Error evaluating the log probability at the initial value."

[230] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[231] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.318, but must be between (0, 1)"

[232] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[233] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[234] "Rejecting initial value:"

[235] " Error evaluating the log probability at the initial value."

[236] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[237] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.299839, but must be between (0, 1)"

[238] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[239] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[240] "Rejecting initial value:"

[241] " Error evaluating the log probability at the initial value."

[242] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[243] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.274129, but must be between (0, 1)"

[244] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[245] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[246] "Rejecting initial value:"

[247] " Error evaluating the log probability at the initial value."

[248] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[249] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0123291, but must be between (0, 1)"

[250] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[251] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[252] "Rejecting initial value:"

[253] " Error evaluating the log probability at the initial value."

[254] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[255] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.78863, but must be between (0, 1)"

[256] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[257] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[258] "Rejecting initial value:"

[259] " Error evaluating the log probability at the initial value."

[260] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[261] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.06626, but must be between (0, 1)"

[262] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[263] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[264] "Rejecting initial value:"

[265] " Error evaluating the log probability at the initial value."

[266] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[267] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.17069, but must be between (0, 1)"

[268] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[269] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[270] "Rejecting initial value:"

[271] " Error evaluating the log probability at the initial value."

[272] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[273] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0699805, but must be between (0, 1)"

[274] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[275] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[276] "Rejecting initial value:"

[277] " Error evaluating the log probability at the initial value."

[278] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[279] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.93048, but must be between (0, 1)"

[280] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[281] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[282] "Rejecting initial value:"

[283] " Error evaluating the log probability at the initial value."

[284] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[285] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.0395, but must be between (0, 1)"

[286] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[287] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[288] "Rejecting initial value:"

[289] " Error evaluating the log probability at the initial value."

[290] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[291] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.71023, but must be between (0, 1)"

[292] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[293] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[294] "Rejecting initial value:"

[295] " Error evaluating the log probability at the initial value."

[296] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[297] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.70424, but must be between (0, 1)"

[298] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[299] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[300] "Rejecting initial value:"

[301] " Error evaluating the log probability at the initial value."

[302] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[303] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.22626, but must be between (0, 1)"

[304] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[305] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[306] "Rejecting initial value:"

[307] " Error evaluating the log probability at the initial value."

[308] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[309] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.68201, but must be between (0, 1)"

[310] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[311] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[312] "Rejecting initial value:"

[313] " Error evaluating the log probability at the initial value."

[314] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[315] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.037525, but must be between (0, 1)"

[316] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[317] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[318] "Rejecting initial value:"

[319] " Error evaluating the log probability at the initial value."

[320] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[321] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.129564, but must be between (0, 1)"

[322] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[323] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[324] "Rejecting initial value:"

[325] " Error evaluating the log probability at the initial value."

[326] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[327] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.3934, but must be between (0, 1)"

[328] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[329] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[330] "Rejecting initial value:"

[331] " Error evaluating the log probability at the initial value."

[332] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[333] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.2274, but must be between (0, 1)"

[334] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[335] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[336] "Rejecting initial value:"

[337] " Error evaluating the log probability at the initial value."

[338] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[339] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.13497, but must be between (0, 1)"

[340] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[341] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[342] "Rejecting initial value:"

[343] " Error evaluating the log probability at the initial value."

[344] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[345] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.306884, but must be between (0, 1)"

[346] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[347] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[348] "Rejecting initial value:"

[349] " Error evaluating the log probability at the initial value."

[350] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[351] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.62236, but must be between (0, 1)"

[352] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[353] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[354] "Rejecting initial value:"

[355] " Error evaluating the log probability at the initial value."

[356] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[357] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.229937, but must be between (0, 1)"

[358] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[359] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[360] "Rejecting initial value:"

[361] " Error evaluating the log probability at the initial value."

[362] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[363] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.12563, but must be between (0, 1)"

[364] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[365] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[366] "Rejecting initial value:"

[367] " Error evaluating the log probability at the initial value."

[368] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[369] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0256329, but must be between (0, 1)"

[370] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[371] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[372] "Rejecting initial value:"

[373] " Error evaluating the log probability at the initial value."

[374] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[375] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0609385, but must be between (0, 1)"

[376] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[377] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[378] "Rejecting initial value:"

[379] " Error evaluating the log probability at the initial value."

[380] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[381] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.08042, but must be between (0, 1)"

[382] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[383] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[384] "Rejecting initial value:"

[385] " Error evaluating the log probability at the initial value."

[386] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[387] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.172817, but must be between (0, 1)"

[388] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[389] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[390] "Rejecting initial value:"

[391] " Error evaluating the log probability at the initial value."

[392] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[393] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.37284, but must be between (0, 1)"

[394] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[395] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[396] "Rejecting initial value:"

[397] " Error evaluating the log probability at the initial value."

[398] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[399] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.83042, but must be between (0, 1)"

[400] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[401] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[402] "Rejecting initial value:"

[403] " Error evaluating the log probability at the initial value."

[404] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[405] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.42001, but must be between (0, 1)"

[406] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[407] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[408] "Rejecting initial value:"

[409] " Error evaluating the log probability at the initial value."

[410] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[411] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.53312, but must be between (0, 1)"

[412] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[413] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[414] "Rejecting initial value:"

[415] " Error evaluating the log probability at the initial value."

[416] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[417] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.91729, but must be between (0, 1)"

[418] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[419] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[420] "Rejecting initial value:"

[421] " Error evaluating the log probability at the initial value."

[422] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[423] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.3365, but must be between (0, 1)"

[424] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[425] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[426] "Rejecting initial value:"

[427] " Error evaluating the log probability at the initial value."

[428] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[429] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.34812, but must be between (0, 1)"

[430] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[431] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[432] "Rejecting initial value:"

[433] " Error evaluating the log probability at the initial value."

[434] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[435] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0728368, but must be between (0, 1)"

[436] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[437] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[438] "Rejecting initial value:"

[439] " Error evaluating the log probability at the initial value."

[440] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[441] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 2.0499, but must be between (0, 1)"

[442] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[443] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[444] "Rejecting initial value:"

[445] " Error evaluating the log probability at the initial value."

[446] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[447] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.45968, but must be between (0, 1)"

[448] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[449] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[450] "Rejecting initial value:"

[451] " Error evaluating the log probability at the initial value."

[452] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[453] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.00981, but must be between (0, 1)"

[454] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[455] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[456] "Rejecting initial value:"

[457] " Error evaluating the log probability at the initial value."

[458] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[459] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.294577, but must be between (0, 1)"

[460] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[461] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[462] "Rejecting initial value:"

[463] " Error evaluating the log probability at the initial value."

[464] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[465] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.02766, but must be between (0, 1)"

[466] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[467] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[468] "Rejecting initial value:"

[469] " Error evaluating the log probability at the initial value."

[470] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[471] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.1845, but must be between (0, 1)"

[472] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[473] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[474] "Rejecting initial value:"

[475] " Error evaluating the log probability at the initial value."

[476] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[477] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 2.3628, but must be between (0, 1)"

[478] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[479] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[480] "Rejecting initial value:"

[481] " Error evaluating the log probability at the initial value."

[482] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[483] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.37675, but must be between (0, 1)"

[484] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[485] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[486] "Rejecting initial value:"

[487] " Error evaluating the log probability at the initial value."

[488] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[489] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.00173473, but must be between (0, 1)"

[490] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[491] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[492] "Rejecting initial value:"

[493] " Error evaluating the log probability at the initial value."

[494] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[495] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.180005, but must be between (0, 1)"

[496] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[497] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[498] "Rejecting initial value:"

[499] " Error evaluating the log probability at the initial value."

[500] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[501] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.06551, but must be between (0, 1)"

[502] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[503] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[504] "Rejecting initial value:"

[505] " Error evaluating the log probability at the initial value."

[506] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[507] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.62884, but must be between (0, 1)"

[508] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[509] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[510] "Rejecting initial value:"

[511] " Error evaluating the log probability at the initial value."

[512] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[513] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.06081, but must be between (0, 1)"

[514] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[515] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[516] "Rejecting initial value:"

[517] " Error evaluating the log probability at the initial value."

[518] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[519] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.22678, but must be between (0, 1)"

[520] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[521] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[522] "Rejecting initial value:"

[523] " Error evaluating the log probability at the initial value."

[524] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[525] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.46573, but must be between (0, 1)"

[526] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[527] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[528] "Rejecting initial value:"

[529] " Error evaluating the log probability at the initial value."

[530] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[531] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0647285, but must be between (0, 1)"

[532] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[533] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[534] "Rejecting initial value:"

[535] " Error evaluating the log probability at the initial value."

[536] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[537] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.195867, but must be between (0, 1)"

[538] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[539] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[540] "Rejecting initial value:"

[541] " Error evaluating the log probability at the initial value."

[542] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[543] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 2.06373, but must be between (0, 1)"

[544] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[545] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[546] "Rejecting initial value:"

[547] " Error evaluating the log probability at the initial value."

[548] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[549] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0360176, but must be between (0, 1)"

[550] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[551] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[552] "Rejecting initial value:"

[553] " Error evaluating the log probability at the initial value."

[554] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[555] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0637329, but must be between (0, 1)"

[556] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[557] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[558] "Rejecting initial value:"

[559] " Error evaluating the log probability at the initial value."

[560] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[561] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0522275, but must be between (0, 1)"

[562] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[563] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[564] "Rejecting initial value:"

[565] " Error evaluating the log probability at the initial value."

[566] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[567] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0721265, but must be between (0, 1)"

[568] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[569] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[570] "Rejecting initial value:"

[571] " Error evaluating the log probability at the initial value."

[572] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[573] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.44547, but must be between (0, 1)"

[574] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[575] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[576] "Rejecting initial value:"

[577] " Error evaluating the log probability at the initial value."

[578] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[579] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.391075, but must be between (0, 1)"

[580] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[581] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[582] "Rejecting initial value:"

[583] " Error evaluating the log probability at the initial value."

[584] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[585] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.1849, but must be between (0, 1)"

[586] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[587] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[588] "Rejecting initial value:"

[589] " Error evaluating the log probability at the initial value."

[590] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[591] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.16468, but must be between (0, 1)"

[592] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[593] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[594] "Rejecting initial value:"

[595] " Error evaluating the log probability at the initial value."

[596] "Informational Message: The current Metropolis proposal is about to be rejected because of the following issue:"

[597] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is 1.06142, but must be between (0, 1)"

[598] "If this warning occurs sporadically, such as for highly constrained variable types like covariance matrices, then the sampler is fine,"

[599] "but if this warning occurs often then your model may be either severely ill-conditioned or misspecified."

[600] "Rejecting initial value:"

[601] " Error evaluating the log probability at the initial value."

[602] "Initialization between (--2, -2) failed after 100 attempts. "

[603] " Try specifying initial values, reducing ranges of constrained values, or reparameterizing the model."

error occurred during calling the sampler; sampling not done

> sessionInfo()

R version 3.2.3 (2015-12-10)

Platform: x86_64-apple-darwin13.4.0 (64-bit)

Running under: OS X 10.11.2 (El Capitan)

locale:

[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:

[1] stats graphics grDevices utils datasets methods base

other attached packages:

[1] rstan_2.9.0 ggplot2_2.0.0 mvtnorm_1.0-3

loaded via a namespace (and not attached):

[1] colorspace_1.2-6 scales_0.3.0 plyr_1.8.3 tools_3.2.3 inline_0.3.14 gtable_0.1.2

[7] gridExtra_2.0.0 Rcpp_0.12.1 codetools_0.2-14 grid_3.2.3 stats4_3.2.3 munsell_0.4.2

I am actually get a little bit overwhelmed with the result and I am wondering whether anyone can help me figure where the problem is?

Thanks!

Sincerely

Jiaqi Zhang

IRT_Mac_OS.R

IRT_Windows_OS.R

Bob Carpenter

unread,

Jan 15, 2016, 1:39:24 PM1/15/16

to stan-...@googlegroups.com

First, we strongly recommend taking the model code out of
the R program and making it standalone so that you can get some
use out of line numbers. I'd also strongly advise indenting
and using standard bracketing.

Also, fewer comments. I always say that comments go stale,
and here's a prime example:

item_diff ~ uniform(0 ,1); # a normal prior for the difficulty

Anyway, there are several problems with this model, the main
one of which is that you need to constrain parameters. If
you really require item_diff in (0,1) uniform, then you need
lower- and upper-bound constraints on item_diff's declaration.
Also, the "guessing" parameter is unconstrained, so you'll get
negative values unless you constrain it or put it through an
(inverse) link function yourself (inv_logit) or constrain it
to fall in (0,1).

- Bob

code_mm<-'data {
int<lower=0> n_student; # n_student: the number of students
int<lower=0> n_item; # n_item: the number of items
int<lower=0,upper=1> y[n_student, n_item]; # the matrix of responses
}

parameters {
vector[n_student] theta; // ability theta for student
vector[n_item] item_diff; // difficulty for item
vector[n_student] guessing; // guessing factor for item
}

model {
// assigning prior distribution for paramaters;
// item_diff is the item difficulty on a uniform(0,1)
item_diff ~ uniform(0 ,1); # a normal prior for the difficulty
// theta is the latent ability with mean 0 and variance 3
theta ~ normal(0,1.732); # a normal prior for the student ability
// guessing follows a prior of uniform(0,1)
// guessing ~ beta(5,23);
guessing ~ uniform(0,1);

// Likelihood:

for(i in 1:n_student)
{
for (j in 1:n_item)
{
y[i,j] ~ bernoulli(guessing[i]*.2+(1-guessing[i])*inv_logit(theta[i] - item_diff[j]));
}
}
}

> --
> You received this message because you are subscribed to the Google Groups "Stan users mailing list" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to stan-users+...@googlegroups.com.
> To post to this group, send email to stan-...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
> <IRT_Mac_OS.R><IRT_Windows_OS.R>

Jiaqi Zhang

unread,

Jan 15, 2016, 2:31:26 PM1/15/16

to Stan users mailing list

Thanks, Bob,

I constrained parameters both item_diff and guessing with lower=0, upper=1. And the model is actually run, I am wondering what other problems you see from the proposed model from your point?

Thanks, your suggestion is really helpful!

Sincerely,

Jiaqi

> [225] "Exception thrown at line 30: stan::math::bernoulli_log: Probability parameter is -0.0209495, but must be b...

Bob Carpenter

unread,

Jan 15, 2016, 2:48:35 PM1/15/16

to stan-...@googlegroups.com

Can you send just the current program? We've coded
most of these IRT models in one way or another.

- Bob

Reply all

Reply to author

Forward

0 new messages