4:57 p.m.
Hi,
I'm working on version 1.8.4 of GRETL.
I've been trying to use programmatically corrgm on two series.
smpl -no-missing F_NFCB E_NFCB
cgm = corrgm(F_NFCB,12,E_NFCB)
but this fails with
gretl version 1.8.4
Current session: 2009-09-07 19:54
? smpl --no-missing F_NFCB E_NFCB
? cgm = corrgm(F_NFCB,12,E_NFCB)
Missing values within sample -- can't do correlogram
Error executing script: halting
cgm = corrgm(F_NFCB,12,E_NFCB)
Any hints or clues what's wrong?
Ofer Cornfeld
9:42 p.m.
On Mon, 7 Sep 2009, Ofer Cornfeld wrote:
I'm working on version 1.8.4 of GRETL.
I've been trying to use programmatically corrgm on two series.
but this fails with
gretl version 1.8.4
Current session: 2009-09-07 19:54
? smpl --no-missing F_NFCB E_NFCB
? cgm = corrgm(F_NFCB,12,E_NFCB)
Missing values within sample -- can't do correlogram
What does the pattern of NAs look like for these two series?
I've now fixed a bug this area that might have tripped gretl
up for some patterns of NAs. The fix is in CVS and the Windows
snapshot.
A workaround that should help with the released gretl 1.8.4 is:
smpl --no-missing F_NFCB E_NFCB
store corrgm.gdt F_NFCB E_NFCB
open corrgm.gdt
cgm = corrgm(F_NFCB,12,E_NFCB)
Allin Cottrell
11:28 p.m.
Dear Allin,
Thanks. I have attached a small script output with the BUG displayed.
You can see the first corrgm is executed with no problem, but the second one
fails with the error message. Hope this will help.
I have used corr function instead in a loop to overcome the bug. The
store/open is too cumbersome within the context of my project.
With kind regards,
Ofer
gretl version 1.8.4
Current session: 2009-09-08 02:25
? smpl full
? print F_NFCB
Full data range: 1975:1 - (n = 130)
NA 506.525 510.300 515.775 517.375 525.100 547.100 564.200
563.600 563.875 588.700 590.950 600.625 603.850 602.850 604.500
607.550 607.300 589.800 590.550 607.000 615.225 621.250 632.975
622.050 612.850 611.850 606.050 599.825 610.900 628.300 645.200
660.650 677.125 691.225 698.000 706.525 711.575 718.550 731.750
734.450 739.350 736.100 738.725 749.075 757.600 773.325 788.925
799.750 811.000 821.925 825.975 841.800 839.850 836.525 840.825
844.625 847.075 856.175 851.950 848.775 842.025 842.025 844.400
847.750 859.975 868.000 868.300 883.250 876.750 890.825 896.725
911.250 924.950 943.725 954.875 974.350 978.200 986.300 1002.03
1010.85 1021.78 1040.30 1058.88 1075.47 1090.22 1106.70 1128.28
1144.10 1155.88 1171.72 1192.10 1205.68 1231.53 1245.78 1255.75
1277.90 1307.13 1314.53 1325.53 1324.97 1321.47 1314.18 1299.40
1289.50 1298.65 1316.35 1324.00 1330.70 1324.65 1336.25 1356.15
1370.47 1386.85 1405.83 1433.33 1436.10 1447.38 1477.78 1475.80
1506.83 1532.40 1535.40 1553.13 1546.83 1547.00 1563.28 1573.33
1580.88 1575.50 1601.97 NA NA NA NA
? print E_NFCB
Full data range: 1975:1 - (n = 131)
55746.5 56840.6 57704.4 58103.0 58523.9 59102.0 59807.1 60389.7
61294.3 61969.7 63048.8 63796.5 64463.2 65113.6 64972.8 65373.1
65906.5 65966.4 64974.2 64798.3 65109.2 66002.4 66662.1 66634.8
66184.9 65807.8 65385.8 65170.4 64812.7 64697.8 65232.7 66634.9
67480.0 68285.3 69536.5 70139.8 70744.4 71120.8 71474.7 71931.6
72470.8 73319.2 73738.6 73912.2 74368.4 74608.5 75238.8 75836.5
76381.8 76749.6 76503.5 77457.6 77837.4 78344.7 79027.4 78977.8
79260.0 80091.8 79968.5 80010.5 79362.2 78506.4 78223.9 78429.8
78263.0 79251.2 79760.8 79421.5 79509.6 79397.5 79792.9 80570.6
81013.9 81911.1 82573.9 83326.1 84519.8 84687.7 84835.8 85164.2
85597.4 85756.3 86332.5 87552.6 87815.0 88145.7 89335.4 90345.8
90912.7 91263.7 91300.5 91589.0 92211.0 93123.6 93244.5 93385.7
94462.6 95624.0 96383.2 96873.9 97380.0 97156.0 96974.4 96766.5
96493.3 96677.8 96410.8 96806.3 96090.9 96886.0 97248.0 96595.2
96926.5 97581.6 98053.5 97895.5 97999.6 98577.0 99571.0 100558.
100830. 101997. 102939. 102026. 102009. 104032. 104102. 102688.
103271. 105037. 104940. NA NA NA NA
? smpl --no-missing F_NFCB E_NFCB
? cgm = corrgm(E_NFCB,12, F_NFCB)
? print cgm
cgm (25 x 1)
0.75301
0.77041
0.78889
0.80709
0.82707
0.84739
0.86597
0.88483
0.90629
0.92745
0.94649
0.96583
0.98638
0.95951
0.93351
0.90686
0.87982
0.85264
0.82574
0.79863
0.77108
0.74419
0.71721
0.69159
0.66717
? cgm = corrgm(F_NFCB,12, E_NFCB)
Missing values within sample -- can't do correlogram
Error executing script: halting
cgm = corrgm(F_NFCB,12, E_NFCB)
-----Original Message-----
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces@lists.wfu.edu] On Behalf Of Allin Cottrell
Sent: Tuesday, September 08, 2009 12:43 AM
To: Gretl list
Subject: Re: [Gretl-users] Corrgm function
On Mon, 7 Sep 2009, Ofer Cornfeld wrote:
I'm working on version 1.8.4 of GRETL.
I've been trying to use programmatically corrgm on two series.
but this fails with
gretl version 1.8.4
Current session: 2009-09-07 19:54
? smpl --no-missing F_NFCB E_NFCB
? cgm = corrgm(F_NFCB,12,E_NFCB)
Missing values within sample -- can't do correlogram
What does the pattern of NAs look like for these two series?
I've now fixed a bug this area that might have tripped gretl
up for some patterns of NAs. The fix is in CVS and the Windows
snapshot.
A workaround that should help with the released gretl 1.8.4 is:
smpl --no-missing F_NFCB E_NFCB
store corrgm.gdt F_NFCB E_NFCB
open corrgm.gdt
cgm = corrgm(F_NFCB,12,E_NFCB)
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
12:28 a.m.
On Tue, 8 Sep 2009, Ofer Cornfeld wrote:
Thanks. I have attached a small script output with the BUG
displayed. You can see the first corrgm is executed with no
problem, but the second one fails with the error message. Hope
this will help.
Thanks. If you could send me the data file you are using that
would be very helpful.
Allin Cottrell
1:02 a.m.
New subject: Corrgm function (after thinking)
On Tue, 8 Sep 2009, Ofer Cornfeld wrote:
Thanks. I have attached a small script output with the BUG
displayed. You can see the first corrgm is executed with no
problem, but the second one fails with the error message. Hope
this will help.
Yes. The short answer is, I'm sorry to say, that the corrgm()
function is nastily broken for datasets including missing values
(for the case where a third argument is given, at any rate) and
should not be trusted at all in that context! I'll put out a
proper fix as soon as I can.
Meanwhile, please note that the xcorrgm command produces correct
results for the cross-correlogram, although of course it doesn't
give you the option of assigning the result to a matrix.
Allin Cottrell
1:31 a.m.
New subject: Corrgm function (after thinking)
On Mon, 7 Sep 2009, Allin Cottrell wrote:
The short answer is, I'm sorry to say, that the corrgm()
function is nastily broken for datasets including missing values
(for the case where a third argument is given, at any rate) and
should not be trusted at all in that context! I'll put out a
proper fix as soon as I can.
OK, the proper fix is now applied in CVS and the Windows snapshot.
Allin Cottrell
11:58 a.m.
New subject: Corrgm function (after thinking)
How do I get a working copy of this Windows snapshot?
With kind regards,
Ofer Cornfeld
-----Original Message-----
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces@lists.wfu.edu] On Behalf Of Allin Cottrell
Sent: Tuesday, September 08, 2009 4:31 AM
To: Gretl list
Subject: Re: [Gretl-users] Corrgm function (after thinking)
On Mon, 7 Sep 2009, Allin Cottrell wrote:
The short answer is, I'm sorry to say, that the corrgm()
function is nastily broken for datasets including missing values
(for the case where a third argument is given, at any rate) and
should not be trusted at all in that context! I'll put out a
proper fix as soon as I can.
OK, the proper fix is now applied in CVS and the Windows snapshot.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
12:46 p.m.
New subject: Corrgm function (after thinking)
From http://ricardo.ecn.wfu.edu/pub/gretl/gretl_install.exe you always
get the most recent version.
Ofer Cornfeld wrote:
How do I get a working copy of this Windows snapshot?
With kind regards,
Ofer Cornfeld
-----Original Message-----
From: gretl-users-bounces(a)lists.wfu.edu
[mailto:gretl-users-bounces@lists.wfu.edu] On Behalf Of Allin Cottrell
Sent: Tuesday, September 08, 2009 4:31 AM
To: Gretl list
Subject: Re: [Gretl-users] Corrgm function (after thinking)
On Mon, 7 Sep 2009, Allin Cottrell wrote:
> The short answer is, I'm sorry to say, that the corrgm()
> function is nastily broken for datasets including missing values
> (for the case where a third argument is given, at any rate) and
> should not be trusted at all in that context! I'll put out a
> proper fix as soon as I can.
OK, the proper fix is now applied in CVS and the Windows snapshot.
Allin Cottrell
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
_______________________________________________
Gretl-users mailing list
Gretl-users(a)lists.wfu.edu
http://lists.wfu.edu/mailman/listinfo/gretl-users
12:47 p.m.
New subject: Corrgm function (after thinking)
On Miércoles, 9 de Septiembre de 2009 13:58:58 Ofer Cornfeld escribió:
How do I get a working copy of this Windows snapshot?
You may get it by clicking on the gretl_install.exe link on the page
http://gretl.sourceforge.net/win32/
--
Ignacio Diaz-Emparanza
DEPARTAMENTO DE ECONOMÍA APLICADA III (ECONOMETRÍA Y ESTADÍSTICA)
UPV/EHU
Avda. Lehendakari Aguirre, 83 | 48015 BILBAO
T.: +34 946013732 | F.: +34 946013754
www.et.bs.ehu.es
5583
days inactive
5585
days old
8 comments
4 participants
participants (4)
-
Allin Cottrell -
chris -
Ignacio Diaz-Emparanza -
Ofer Cornfeld