Hi Steve,
So here is a sample data set from one of my tsplot directories. Columns 1-4 are from
tsplot_zstat3.txt and columns 5-8 are from tsplot_zstat4.txt. I included only the first 10 rows.
1.0e+04 *
1.2794 1.2813 1.2808 1.2799 1.4178 1.4080 1.4150 1.4108
1.2863 1.2813 1.2810 1.2866 1.4172 1.4080 1.4139 1.4114
1.2739 1.2813 1.2810 1.2742 1.4047 1.4080 1.4134 1.3993
1.2839 1.2813 1.2807 1.2846 1.4178 1.4080 1.4136 1.4123
1.2695 1.2814 1.2808 1.2701 1.4156 1.4080 1.4141 1.4095
1.2837 1.2813 1.2820 1.2831 1.4141 1.4080 1.4117 1.4103
1.2919 1.2811 1.2811 1.2919 1.4170 1.4079 1.4082 1.4168
1.2745 1.2820 1.2798 1.2766 1.4095 1.4079 1.4050 1.4125
1.2942 1.2834 1.2807 1.2969 1.4039 1.4079 1.4041 1.4077
1.2752 1.2834 1.2812 1.2774 1.4119 1.4079 1.4052 1.4145
As I understand it, the columns correspond to
[fulldata, copepartialmodelfit, fullmodelfit, reduceddata].
As you can see, the fulldata (columns 1 & 4) are very different, as is the fullmodelfit (columns 3 &
7). Furthermore, I calculated the residuals (reduceddata-copepartialmodelfit) from
tsplot_zstat3.txt and tsplot_zstat4.txt:
-13.5800 27.8900
53.1200 33.3800
-70.7800 -86.7600
32.1100 42.8100
-113.1000 15.4500
17.2700 23.7800
107.9000 88.5300
-53.5200 45.8000
134.9500 -1.7500
-59.5300 66.1800
Again, they are very different. The fulldata, fullmodelfit, and residual should be the same across
EVs. This pattern is consistent across all instances of tsplot that I've run, and it doesn't matter
whether the mask is binarized or not. I'm clearly missing something. thanks,
jack
On Mon, 17 Oct 2005 16:04:27 +0100, Stephen Smith <[log in to unmask]> wrote:
>Hi Jack - if you run tsplot with the "-m mask" option and the -n tsplot
>option and are not doing any other thresholding then yes the timeseries
>data should be the same.....
>
>Cheers, Steve.
>
>
>On Mon, 10 Oct 2005, Jack Grinband wrote:
>
>> Hi Steve,
>> So, by default, all the stats in featquery are calculated using a binary mask but all the plots are
>> calculated using a weighted mask. Is this right?
>>
>> I ran tsplot using -n, but I'm still getting different values for "data", "full model fit", and the
>> residuals across EVs. It seems to me that none of these should change. Is there a reason why
>> they're different?
>> thanks,
>>
>> jack
|