12:03 p.m.
Hello,
Using libgretl 1.8.7, with heteroscedacity function, i got the error :
"Weight variable is all zeros, aborting regression"
but when i add one observations it works or some observations added later it
does the same error, more observations later same it works and so on.
I give you the data with errors, it's a time serie with weekly observations,
i use this before calling the hsk_func :
listlag = laggenr_from_to(1, 1, 75, &m_Z, m_datainfo, &err);
and the m_Z contains at the beginning :
m_Z[1][0] = 1431225.000000;
m_Z[1][1] = 207958.000000;
m_Z[1][2] = 1328798.000000;
m_Z[1][3] = 1412733.000000;
m_Z[1][4] = 1766147.000000;
m_Z[1][5] = 1635199.000000;
m_Z[1][6] = 1565249.000000;
m_Z[1][7] = 949787.000000;
m_Z[1][8] = 1130386.000000;
m_Z[1][9] = 834164.000000;
m_Z[1][10] = 2080654.000000;
m_Z[1][11] = 1153485.000000;
m_Z[1][12] = 857538.000000;
m_Z[1][13] = 1363286.000000;
m_Z[1][14] = 2051956.000000;
m_Z[1][15] = 660288.000000;
m_Z[1][16] = 664975.000000;
m_Z[1][17] = 1831966.000000;
m_Z[1][18] = 1516991.000000;
m_Z[1][19] = 1284537.000000;
m_Z[1][20] = 978315.000000;
m_Z[1][21] = 611010.000000;
m_Z[1][22] = 86147.000000;
m_Z[1][23] = 1004510.000000;
m_Z[1][24] = 150055.000000;
m_Z[1][25] = 632921.000000;
m_Z[1][26] = 182371.000000;
m_Z[1][27] = 1269861.000000;
m_Z[1][28] = 599457.000000;
m_Z[1][29] = 989400.000000;
m_Z[1][30] = 859726.000000;
m_Z[1][31] = 1501113.000000;
m_Z[1][32] = 249601.000000;
m_Z[1][33] = 383646.000000;
m_Z[1][34] = 311465.000000;
m_Z[1][35] = 1591434.000000;
m_Z[1][36] = 817778.000000;
m_Z[1][37] = 906471.000000;
m_Z[1][38] = 1957765.000000;
m_Z[1][39] = 1024629.000000;
m_Z[1][40] = 1293944.000000;
m_Z[1][41] = 920355.000000;
m_Z[1][42] = 1956842.000000;
m_Z[1][43] = 1560762.000000;
m_Z[1][44] = 1931361.000000;
m_Z[1][45] = 1326024.000000;
m_Z[1][46] = 1838782.000000;
m_Z[1][47] = 205036.000000;
m_Z[1][48] = 1168592.000000;
m_Z[1][49] = 1081148.000000;
m_Z[1][50] = 1741196.000000;
m_Z[1][51] = 798779.000000;
m_Z[1][52] = 103289.000000;
m_Z[1][53] = 456865.000000;
m_Z[1][54] = 543444.000000;
m_Z[1][55] = 1407659.000000;
m_Z[1][56] = 1519187.000000;
m_Z[1][57] = 375135.000000;
m_Z[1][58] = 852668.000000;
m_Z[1][59] = 1772748.000000;
m_Z[1][60] = 134335.000000;
m_Z[1][61] = 687954.000000;
m_Z[1][62] = 1827437.000000;
m_Z[1][63] = 762331.000000;
m_Z[1][64] = 603293.000000;
m_Z[1][65] = 792405.000000;
m_Z[1][66] = 1822889.000000;
m_Z[1][67] = 28162.000000;
m_Z[1][68] = 119814.000000;
m_Z[1][69] = 1860604.000000;
m_Z[1][70] = 1121620.000000;
m_Z[1][71] = 1684701.000000;
m_Z[1][72] = 1077879.000000;
m_Z[1][73] = 1926145.000000;
m_Z[1][74] = 724391.000000;
m_Z[1][75] = 2115458.000000;
m_Z[1][76] = 7795.000000;
m_Z[1][77] = 1047352.000000;
m_Z[1][78] = 840371.000000;
m_Z[1][79] = 1476913.000000;
m_Z[1][80] = 1511875.000000;
m_Z[1][81] = 176943.000000;
m_Z[1][82] = 133643.000000;
m_Z[1][83] = 887549.000000;
m_Z[1][84] = 1812511.000000;
m_Z[1][85] = 17846.000000;
m_Z[1][86] = 603354.000000;
m_Z[1][87] = 1126207.000000;
m_Z[1][88] = 1232769.000000;
m_Z[1][89] = 196866.000000;
m_Z[1][90] = 681670.000000;
m_Z[1][91] = 664312.000000;
m_Z[1][92] = 291783.000000;
m_Z[1][93] = 1872033.000000;
m_Z[1][94] = 1054658.000000;
m_Z[1][95] = 689063.000000;
m_Z[1][96] = 1023127.000000;
m_Z[1][97] = 755405.000000;
m_Z[1][98] = 1742680.000000;
m_Z[1][99] = 1085702.000000;
m_Z[1][100] = 967574.000000;
m_Z[1][101] = 57139.000000;
m_Z[1][102] = 1967121.000000;
m_Z[1][103] = 155571.000000;
m_Z[1][104] = 22136.000000;
m_Z[1][105] = 876667.000000;
m_Z[1][106] = 1834178.000000;
m_Z[1][107] = 609363.000000;
m_Z[1][108] = 2082755.000000;
m_Z[1][109] = 195996.000000;
m_Z[1][110] = 1858837.000000;
m_Z[1][111] = 191434.000000;
m_Z[1][112] = 56269.000000;
m_Z[1][113] = 2036749.000000;
m_Z[1][114] = 504633.000000;
m_Z[1][115] = 1218606.000000;
m_Z[1][116] = 117861.000000;
m_Z[1][117] = 97979.000000;
m_Z[1][118] = 1988992.000000;
m_Z[1][119] = 2057725.000000;
m_Z[1][120] = 361376.000000;
m_Z[1][121] = 1357918.000000;
m_Z[1][122] = 928988.000000;
m_Z[1][123] = 543378.000000;
m_Z[1][124] = 1518102.000000;
m_Z[1][125] = 1585398.000000;
m_Z[1][126] = 957590.000000;
m_Z[1][127] = 84222.000000;
m_Z[1][128] = 1572125.000000;
m_Z[1][129] = 1017730.000000;
m_Z[1][130] = 1918768.000000;
m_Z[1][131] = 71380.000000;
m_Z[1][132] = 214801.000000;
m_Z[1][133] = 190478.000000;
m_Z[1][134] = 1240358.000000;
m_Z[1][135] = 2002441.000000;
m_Z[1][136] = 2077945.000000;
m_Z[1][137] = 1095170.000000;
m_Z[1][138] = 908149.000000;
m_Z[1][139] = 1424074.000000;
m_Z[1][140] = 1899931.000000;
m_Z[1][141] = 54489.000000;
m_Z[1][142] = 1944286.000000;
m_Z[1][143] = 931712.000000;
m_Z[1][144] = 964216.000000;
m_Z[1][145] = 817239.000000;
m_Z[1][146] = 1491468.000000;
m_Z[1][147] = 1804150.000000;
m_Z[1][148] = 754704.000000;
m_Z[1][149] = 1526318.000000;
m_Z[1][150] = 1261455.000000;
m_Z[1][151] = 1658306.000000;
m_Z[1][152] = 1138944.000000;
m_Z[1][153] = 740934.000000;
m_Z[1][154] = 872778.000000;
m_Z[1][155] = 931733.000000;
m_Z[1][156] = 445110.000000;
m_Z[1][157] = 156769.000000;
m_Z[1][158] = 1101183.000000;
m_Z[1][159] = 274844.000000;
m_Z[1][160] = 371715.000000;
m_Z[1][161] = 2082148.000000;
m_Z[1][162] = 1518744.000000;
m_Z[1][163] = 522768.000000;
m_Z[1][164] = 361869.000000;
m_Z[1][165] = 1427257.000000;
m_Z[1][166] = 1729193.000000;
m_Z[1][167] = 583271.000000;
m_Z[1][168] = 1734855.000000;
m_Z[1][169] = 2019046.000000;
m_Z[1][170] = 2051469.000000;
m_Z[1][171] = 10471.000000;
m_Z[1][172] = 671369.000000;
m_Z[1][173] = 1689988.000000;
m_Z[1][174] = 813742.000000;
m_Z[1][175] = 985851.000000;
m_Z[1][176] = 1701341.000000;
m_Z[1][177] = 1499903.000000;
m_Z[1][178] = 327022.000000;
m_Z[1][179] = 110924.000000;
m_Z[1][180] = 1472301.000000;
m_Z[1][181] = 1494097.000000;
m_Z[1][182] = 2073332.000000;
m_Z[1][183] = 686139.000000;
m_Z[1][184] = 259252.000000;
m_Z[1][185] = 1318475.000000;
m_Z[1][186] = 696237.000000;
m_Z[1][187] = 1232458.000000;
m_Z[1][188] = 1335898.000000;
m_Z[1][189] = 636697.000000;
m_Z[1][190] = 635087.000000;
m_Z[1][191] = 1913213.000000;
m_Z[1][192] = 838686.000000;
m_Z[1][193] = 1514290.000000;
m_Z[1][194] = 813316.000000;
m_Z[1][195] = 947732.000000;
m_Z[1][196] = 1333502.000000;
m_Z[1][197] = 322433.000000;
m_Z[1][198] = 871298.000000;
m_Z[1][199] = 368299.000000;
m_Z[1][200] = 580202.000000;
m_Z[1][201] = 638802.000000;
m_Z[1][202] = 1697618.000000;
m_Z[1][203] = 1467552.000000;
m_Z[1][204] = 5388.000000;
m_Z[1][205] = 432330.000000;
m_Z[1][206] = 1685650.000000;
m_Z[1][207] = 2100829.000000;
m_Z[1][208] = 572783.000000;
m_Z[1][209] = 398972.000000;
m_Z[1][210] = 39131.000000;
m_Z[1][211] = 2089804.000000;
m_Z[1][212] = 1262827.000000;
m_Z[1][213] = 337849.000000;
m_Z[1][214] = 939217.000000;
m_Z[1][215] = 1390430.000000;
m_Z[1][216] = 1862614.000000;
m_Z[1][217] = 187631.000000;
m_Z[1][218] = 126734.000000;
m_Z[1][219] = 1060630.000000;
m_Z[1][220] = 386496.000000;
m_Z[1][221] = 1626442.000000;
m_Z[1][222] = 890342.000000;
m_Z[1][223] = 887525.000000;
m_Z[1][224] = 2007208.000000;
m_Z[1][225] = 76483.000000;
with this it doesn't work but when i add one observation :
m_Z[1][226] = 58168.000000;
it works...
is it normal ? or is it a bug ?
thanks.
4:24 p.m.
On Wed, 28 Apr 2010, denis joubert wrote:
Using libgretl 1.8.7, with heteroscedacity function, i got the error
:
"Weight variable is all zeros, aborting regression"
but when i add one observations it works or some observations added later it
does the same error, more observations later same it works and
so on...
The test libgretl was using for "weights all zero" was not really
appropriate: it was that the sum of squares of the weight series
equals machine zero. I've now modified this in CVS and your
example runs. (The weights are extremely small, but not literally
zero.)
However, the example is rather odd and this may explain gretl's
complaints. The auxiliary regression for hsk estimation produces a
perfect fit with the original dataset (T = 226), since you have
151 usable observations and 151 regressors (75 lags, 75 squares of
lags, plus constant). Adding another observation gives one degree
of freedom to the auxiliary regression, which seems to help
matters.
Allin Cottrell
7:27 a.m.
2010/5/3 Allin Cottrell <cottrell(a)wfu.edu>:
On Wed, 28 Apr 2010, denis joubert wrote:
> Using libgretl 1.8.7, with heteroscedacity function, i got the error :
> "Weight variable is all zeros, aborting regression"
> but when i add one observations it works or some observations added later it
> does the same error, more observations later same it works and
> so on...
The test libgretl was using for "weights all zero" was not really
appropriate: it was that the sum of squares of the weight series
equals machine zero. I've now modified this in CVS and your
example runs. (The weights are extremely small, but not literally
zero.)
thanks, i will get the cvs version :)
However, the example is rather odd and this may explain gretl's
complaints. The auxiliary regression for hsk estimation produces a
perfect fit with the original dataset (T = 226), since you have
151 usable observations and 151 regressors (75 lags, 75 squares of
lags, plus constant). Adding another observation gives one degree
of freedom to the auxiliary regression, which seems to help
matters.
the 151 observations with 151 regressors is only an example of the
problem i got, with 210 observations or more i got the same problem (I
added 1 observations at a time, and more I add observations more the
problem shows up)
does it mean i don't need to use hsk and should use only the auxiliary
regression function in order to predict next values ?
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
2 p.m.
On Tue, 4 May 2010, denis joubert wrote:
2010/5/3 Allin Cottrell <cottrell(a)wfu.edu>:
> The test libgretl was using for "weights all zero" was not really
> appropriate: it was that the sum of squares of the weight series
> equals machine zero. I've now modified this in CVS and your
> example runs. (The weights are extremely small, but not literally
> zero.)
thanks, i will get the cvs version :)
>
> However, the example is rather odd and this may explain gretl's
> complaints. The auxiliary regression for hsk estimation produces a
> perfect fit with the original dataset (T = 226), since you have
> 151 usable observations and 151 regressors (75 lags, 75 squares of
> lags, plus constant). Adding another observation gives one degree
> of freedom to the auxiliary regression, which seems to help
> matters.
the 151 observations with 151 regressors is only an example of the
problem i got, with 210 observations or more i got the same problem (I
added 1 observations at a time, and more I add observations more the
problem shows up)
does it mean i don't need to use hsk and should use only the auxiliary
regression function in order to predict next values ?
No, the auxiliary regression is predicting the error variance, not
the dependent variable. I wonder, though, why you reckon you need
75 lags in this model (other than, perhaps, to stress-test
libgretl!).
Allin Cottrell
4:01 p.m.
2010/5/4 Allin Cottrell <cottrell(a)wfu.edu>:
On Tue, 4 May 2010, denis joubert wrote:
> 2010/5/3 Allin Cottrell <cottrell(a)wfu.edu>:
> > The test libgretl was using for "weights all zero" was not really
> > appropriate: it was that the sum of squares of the weight series
> > equals machine zero. I've now modified this in CVS and your
> > example runs. (The weights are extremely small, but not literally
> > zero.)
>
> thanks, i will get the cvs version :)
> >
> > However, the example is rather odd and this may explain gretl's
> > complaints. The auxiliary regression for hsk estimation produces a
> > perfect fit with the original dataset (T = 226), since you have
> > 151 usable observations and 151 regressors (75 lags, 75 squares of
> > lags, plus constant). Adding another observation gives one degree
> > of freedom to the auxiliary regression, which seems to help
> > matters.
>
> the 151 observations with 151 regressors is only an example of the
> problem i got, with 210 observations or more i got the same problem (I
> added 1 observations at a time, and more I add observations more the
> problem shows up)
>
> does it mean i don't need to use hsk and should use only the auxiliary
> regression function in order to predict next values ?
No, the auxiliary regression is predicting the error variance, not
the dependent variable. I wonder, though, why you reckon you need
75 lags in this model (other than, perhaps, to stress-test
libgretl!).
75 was a bit to stress libgretl (i stressed libgretl with 100 lags the
first time i used it).
But the datas are very hard to predict correctly, i was trying
predicting using 100 lags with gretl gui,
the datas are made from others and can be partially reversed. (the
informations alone are very hard to be correctly predicted)
But i successfully predicted some informations about the informations
which made the datas correctly This attemps was not serious.
So now i do the tries as a clockwork (but i tried 75 in order to
reduce the amount of informations if it does work too i will reduce
the number of lags). this try is connected this time to machine
learning algorithm in order to recover the basis informations (the
first time it was made with my brain and hands).
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
10:47 p.m.
Hello Allin,
I think there is a bug into laggenr_from_to or i don't know to use it.
when you call it twice on 2 different id of variable(already stored
into Z), it does restart the position of the returned list.
For example :
listlag1 = laggenr_from_to(1, 1, 50, &m_Z, m_datainfo, &err);
listlag2 = laggenr_from_to(2, 1, 50, &m_Z, m_datainfo, &err);
return the same numbers into listlag1 and listlag2.
I tried to fix myself the problem but i'm going to sleep. so i stop
not far further than get_transform() (called by laggenr, called by
laggenr_from_to)
2010/5/4 denis joubert <denis.joubert(a)gmail.com>:
2010/5/4 Allin Cottrell <cottrell(a)wfu.edu>:
>
> On Tue, 4 May 2010, denis joubert wrote:
>
>> 2010/5/3 Allin Cottrell <cottrell(a)wfu.edu>:
>> > The test libgretl was using for "weights all zero" was not really
>> > appropriate: it was that the sum of squares of the weight series
>> > equals machine zero. I've now modified this in CVS and your
>> > example runs. (The weights are extremely small, but not literally
>> > zero.)
>>
>> thanks, i will get the cvs version :)
>> >
>> > However, the example is rather odd and this may explain gretl's
>> > complaints. The auxiliary regression for hsk estimation produces a
>> > perfect fit with the original dataset (T = 226), since you have
>> > 151 usable observations and 151 regressors (75 lags, 75 squares of
>> > lags, plus constant). Adding another observation gives one degree
>> > of freedom to the auxiliary regression, which seems to help
>> > matters.
>>
>> the 151 observations with 151 regressors is only an example of the
>> problem i got, with 210 observations or more i got the same problem (I
>> added 1 observations at a time, and more I add observations more the
>> problem shows up)
>>
>> does it mean i don't need to use hsk and should use only the auxiliary
>> regression function in order to predict next values ?
>
> No, the auxiliary regression is predicting the error variance, not
> the dependent variable. I wonder, though, why you reckon you need
> 75 lags in this model (other than, perhaps, to stress-test
> libgretl!).
75 was a bit to stress libgretl (i stressed libgretl with 100 lags the
first time i used it).
But the datas are very hard to predict correctly, i was trying
predicting using 100 lags with gretl gui,
the datas are made from others and can be partially reversed. (the
informations alone are very hard to be correctly predicted)
But i successfully predicted some informations about the informations
which made the datas correctly This attemps was not serious.
So now i do the tries as a clockwork (but i tried 75 in order to
reduce the amount of informations if it does work too i will reduce
the number of lags). this try is connected this time to machine
learning algorithm in order to recover the basis informations (the
first time it was made with my brain and hands).
>
> Allin Cottrell
> _______________________________________________
> Gretl-users mailing list
> Gretl-users(a)lists.wfu.edu
> http://lists.wfu.edu/mailman/listinfo/gretl-users
>
12:15 a.m.
On Wed, 5 May 2010, denis joubert wrote:
I think there is a bug into laggenr_from_to or i don't know to
use it.
when you call it twice on 2 different id of variable(already stored
into Z), it does restart the position of the returned list.
For example :
listlag1 = laggenr_from_to(1, 1, 50, &m_Z, m_datainfo, &err);
listlag2 = laggenr_from_to(2, 1, 50, &m_Z, m_datainfo, &err);
return the same numbers into listlag1 and listlag2.
You will have to show me a small but complete program that
produces unexpected results to make me believe there is a bug. The
following program produces two distinct lists: 3 lags of the first
variable, then 3 lags of the second.
#include <gretl/libgretl.h>
int main (void)
{
DATAINFO *m_datainfo;
double **m_Z;
int *laglist1, *laglist2;
PRN *prn;
int err = 0;
libgretl_init();
prn = gretl_print_new(GRETL_PRINT_STDOUT, NULL);
m_datainfo = create_new_dataset(&m_Z, 3, 10, 0);
strcpy(m_datainfo->varname[1], "y1");
strcpy(m_datainfo->varname[2], "y2");
m_Z[1][0] = 1431225;
m_Z[1][1] = 207958;
m_Z[1][2] = 1328798;
m_Z[1][3] = 1412733;
m_Z[1][4] = 1766147;
m_Z[1][5] = 1635199;
m_Z[1][6] = 1565249;
m_Z[1][7] = 949787;
m_Z[1][8] = 1130386;
m_Z[1][9] = 834164;
m_Z[2][0] = 2080654;
m_Z[2][1] = 1153485;
m_Z[2][2] = 857538;
m_Z[2][3] = 1363286;
m_Z[2][4] = 2051956;
m_Z[2][5] = 660288;
m_Z[2][6] = 664975;
m_Z[2][7] = 1831966;
m_Z[2][8] = 1516991;
m_Z[2][9] = 1284537;
laglist1 = laggenr_from_to(1, 1, 3, &m_Z, m_datainfo, &err);
laglist2 = laggenr_from_to(2, 1, 3, &m_Z, m_datainfo, &err);
printlist(laglist1, "list of lags for var 1");
printlist(laglist2, "list of lags for var 2");
free(laglist1);
free(laglist2);
destroy_dataset(m_Z, m_datainfo);
gretl_print_destroy(prn);
libgretl_cleanup();
return 0;
}
Allin Cottrell
7:35 a.m.
Hello Allin,
I found my bug, if m_datainfo->varname[2] is not set, it give back the
same list at laglist2.
example
replace in your example:
strcpy(m_datainfo->varname[1], "y1");
strcpy(m_datainfo->varname[2], "y2");
by
strcpy(m_datainfo->varname[0], "y1");
strcpy(m_datainfo->varname[1], "y2");
and you'll get my problem.
I'm not sure it's a bug, but it's a big trap.
thanks
2010/5/5 Allin Cottrell <cottrell(a)wfu.edu>:
On Wed, 5 May 2010, denis joubert wrote:
> I think there is a bug into laggenr_from_to or i don't know to use it.
> when you call it twice on 2 different id of variable(already stored
> into Z), it does restart the position of the returned list.
> For example :
> listlag1 = laggenr_from_to(1, 1, 50, &m_Z, m_datainfo, &err);
> listlag2 = laggenr_from_to(2, 1, 50, &m_Z, m_datainfo, &err);
> return the same numbers into listlag1 and listlag2.
You will have to show me a small but complete program that
produces unexpected results to make me believe there is a bug. The
following program produces two distinct lists: 3 lags of the first
variable, then 3 lags of the second.
#include <gretl/libgretl.h>
int main (void)
{
DATAINFO *m_datainfo;
double **m_Z;
int *laglist1, *laglist2;
PRN *prn;
int err = 0;
libgretl_init();
prn = gretl_print_new(GRETL_PRINT_STDOUT, NULL);
m_datainfo = create_new_dataset(&m_Z, 3, 10, 0);
strcpy(m_datainfo->varname[1], "y1");
strcpy(m_datainfo->varname[2], "y2");
m_Z[1][0] = 1431225;
m_Z[1][1] = 207958;
m_Z[1][2] = 1328798;
m_Z[1][3] = 1412733;
m_Z[1][4] = 1766147;
m_Z[1][5] = 1635199;
m_Z[1][6] = 1565249;
m_Z[1][7] = 949787;
m_Z[1][8] = 1130386;
m_Z[1][9] = 834164;
m_Z[2][0] = 2080654;
m_Z[2][1] = 1153485;
m_Z[2][2] = 857538;
m_Z[2][3] = 1363286;
m_Z[2][4] = 2051956;
m_Z[2][5] = 660288;
m_Z[2][6] = 664975;
m_Z[2][7] = 1831966;
m_Z[2][8] = 1516991;
m_Z[2][9] = 1284537;
laglist1 = laggenr_from_to(1, 1, 3, &m_Z, m_datainfo, &err);
laglist2 = laggenr_from_to(2, 1, 3, &m_Z, m_datainfo, &err);
printlist(laglist1, "list of lags for var 1");
printlist(laglist2, "list of lags for var 2");
free(laglist1);
free(laglist2);
destroy_dataset(m_Z, m_datainfo);
gretl_print_destroy(prn);
libgretl_cleanup();
return 0;
}
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
5099
days inactive
5106
days old
7 comments
2 participants
participants (2)
-
Allin Cottrell -
denis joubert