EXPORT HELP

[Gambler]

I downloaded CD 11/4.
in htr program i get pl4 nk value--68--pl5 nk value--67
then i export hx4 over to a table in access2003 and import pl4&5
--pl4 nk value is 71.105---pl5 nk value is 67.1157
pl4 nk value is different.
what am i doing wrong?
thanks
ed

[Rick]

You need to be more specific.

What race and which horses.

I'll download and take a look but I need to know where you are getting the different numbers.

[Gambler]

Sorry Rick
11/4 CD race1 horse #1 Dixie Dream
thanks rick

[Rick]

You didn't do anything wrong.

I got the same results.

km will have to take a look at it.

[Gambler]

Once again thanks a million Rick. I thought i was doing something wrong.
Thanks Rick
ed

[Rick]

Thanks for bringing it up.

That is the only way that problems in the program get fixed is when someone finds them.

Now km gets to figure out what is wrong.

[Donnie]

Guys-
won't the K number change as per the paceline chosen? AQU today...R1...Horse #1 Annabelle....PL5 K=113.....PL1 K=111.

[Rick]

Ed was pointing out that the even though the numbers change that the numbers in PL4 were different from what was being displayed in the program and what was in the export.

#1 in the 1st race in PL4

In HTR2 screen (K) = 068
In Export (K) = 71.1057

I have looked at seven of the races so far and have not found any difference in the (K) rating under PL4.

[km]

I'm looking right at it in HTR2 and Excel (nK) for CD, Race-1, Nov, 4, Dixie Dreamer

Krating . Screen (Export)

PL-4 = 76 (76.2657)
PL-5 = 75 (74.9735)

Might you have exported without the scratches? There were 4 of them in here.

It is not possible for the Export and the screen not to match, they are the exact same variable. In the past when this problem was brought up, the usual reason was either scratches or track surface change (Turf to Wet, for example).

[Rick]

I know that I exported without the scratches and looked at the K ratings in HTR2 without the scratches and the numbers were not the same.

I just downloaded the racefile without the results or charts.

[km]

Sorry Rick and Ed, I am unable to repeat your output.

MikeDee or Donnie, will you please verify the (K) matches in the export (HX4) in Access.

re:
CD, Nov 4, 2006, Race-1, Horse #1, Dixie

1) I created a new temp folder and downloaded CD, Nov 4 into that folder WITHOUT the results or charts. Just the RAcefile.

2) Looking right at the horse in the PRG screen: PL5 (K) = 67; PL4 (K) = 68

3) Exported HX4 with PL-5, then PL-4

4) Dixie (#1) shows

(nK) = 67.1157 for PL-5
(nK) = 68.4457 for PL-4

It matches the screen.
I'm using MS Excel to view the output.

This was the Br Cup file, an early file. Is it possible that has something to do with this?
Can you find your bug in any other output from another track or date?
Want to help, but just can't duplicate the problem from my end.

[njcurveball]

Ken, et al.

I am not at my laptop, but wanted to offer that there may be two separate files involved here. This sounds familiar to last year. They posted a file on Thursday and another on Friday.

Before gettng the new one on Friday, I deleted the Thursday version. Is there a potential for this to happen if they tried to download everything on Friday but then had a Thursday export? Or some permutation of that?

Just wanted to offer a fresh view.

Jim

[Rick]

I downloaded the file yesterday.

The 2nd time I did the export in PL4 it came out right.

Strange shit.

[Rick]

The 3rd time I did the export I did it like I did the first time.

I export PL5 and then export PL4 to the same file.

Now the results are the same as the first time 71.1057

[MikeDee]

I just did a comparison of the first 5 races at beu for nov 7th. No results added, no scratches, just the main b files.

I exported PL 5 data into my db

I compared the k value on the km screen with the k value in the hx4 export.

In every case the rounded the hx4 value matches the k value on the km screen.

I then cleared the data in the kx4 table.

I then exported PL4 data into my hx4 access db.

I them compared the rounded kvalue in the hx4 table to the k value on the km screen for pl 4.

Again each and every value is the same.

The data is getting to cleveland OK :)

[Rick]

Ed,

2 suggestions if you are going to export both PL5 and PL4.

Either export PL4 first or export PL5, close HTR2 and then reopen and export PL4.

If there is a bug, it only seems to appear in the 1st horse in the 1st race when you happen to export PL5 and then PL4.

That is the only way I was able to duplicate it.

It may have only been that one race.

I will do some testing and see if I can duplicate the same thing on a different track or date.

[km]

I'll check it out, i think Rick has put his finger on it. Perhaps the memory isn't clearing before the 2nd PL is chosen and exported.

Not a problem if using just one PL choice or re-starting the export as Rick mentioned.

The bug only seems to be the 1st horse in the 1st race after switching PL modes, so that should lead me in the right direction to fix the problem.

thank you, very obscure situation, nice catch Ed

[Gambler]

CD 11/4 race1 #3 fiery htr has pl4 kline as 64.5 and exporter has it as 74

[Gambler]

I just tried your way Rick and they both htr and export came ot the same. Note
that was the third horse.

[Rick]

I got 64.5.

Did you export PL4 first or by itself?

[Rick]

Yes, but if the first horse changes it could change others.

km is suppose to take a look at it next week and see if he can figure out what the problem is.

Until he fixes it, don't mix PL4 exports in with other exports or always export PL4 first.

For now that seems the only way around the problem.

[Gambler]

Thanks Rick

[Gambler]

Hi Rick
Do you know if Ken had time to look at this problem. I use both PL4 & 5 in my program and havent been able to find a way to export the data correctly.
Thanks
ed

[Rick]

You will have to ask Ken. Which I guess you just did. :)

Are you saying that doing the PL4 export first or by itself doesn't work?

I only tried it once and it appeared to work for me.

I would just make sure and try doing the PL4 export the first thing after you open HTR2.

Other than that you got to wait for km.

[km]

I can't re create the problem Ed. I'll experiment some more and try to figure out why PL-4 would not work in a consecutive manner, but it makes no logical sense to me at the moment. The code does not appear to have a bug as all the methods work the same.

In the meantime, use the technique Rick suggested, export with one PL mode, then exit the screen and return and export the next one.

[Donnie]

Ken-
Is Ed exporting from a cutom program? Possibly something wrong in his version? Dunno, just a guess.....

[Gambler]

I agree with rick, strange shit. Sometimes it good and sometimes it bad.
I downloaded 11/12 CD. Exported PL4. Completely got put of HTR2 [ im not using my custom verson]. Then i got back in and exported PL5. Race 2 HUMBLE K #s are correcf for both pacelines. The Kline number for pl4 is wrong. For some horses they are correct and for some there not. One has to sometime look at several horses and compare their exported paceline against HTR2 before they will find differences. i cant fine a way to export so the numbers are correct all the time. Any advise.
Im using windows xp and access 2003. Im looking in the access table to get my numbers

[km]

Let's to figure this out methodically. Please confirm the following =

1) Using the "Oct 2, 2006" version of HTR2.EXE or HTR2006.EXE

2) Downloaded bare file from CD Nov 12,2006; No Results, No Scratches, No Surface Change.

3) RAce-2, this is the race Ed alludes to. K Lines on the SCREEN (not export), please verify below.

PL-5 from the top down
#8 = 2.9
#6 = 3.9
#3 = 5.1
#2 = 6.0
#1 = 11.2
#7 = 13.0
#4 = 18.6
#5 = 31.7

PL-4 from the top down
#6 = 3.1
#8 = 4.0
#3 = 4.9
#2 = 6.3
#1 = 9.6
#7 = 12.6
#4 = 18.0
#5 = 30.7

Notice that the rank order changes when the PL mode is switched.
Do you concur with the above on your screen? If so, then i'll export next.

[km]

Now here is the exported KLInes, i'm looking at the raw text and then Excel.

I exported back to back without leaving the program. PL-5, then PL-4.

PL-5
#1 = 11.2499 (11.2)
#2 = 6 (6.0)
#3 = 5.125 (5.1)
#4 = 18.6002 (18.6)
#5 = 31.6669 (31.7)
#5 = 3.9 (3.9)
#7 = 12.9999 (13.0)
#8 = 2.92 (2.9)

EXACT Match

PL-4
#1 = 9.5555 (9.6)
#2 = 6.3077 (6.3)
#3 = 4.9375 (4.9)
#4 = 17.9998 (18.0)
#5 = 30.6666 (30.7)
#6 = 3.1304 (3.1)
#7 = 12.5715 (12.6)
#8 = 4 (4.0)

EXACT Match

Is it possible you mis interpreted the 4 digit decimal when comparing to the screen? For example #4, is 17.9998, but 18.0 on the screen when rounded.

I cannot fix what is not broken. I'm looking at the text and the Excel and they match in all cases. Is it possible that something in Access could be mixing up the numbers? I doubt that, but anything is possible, so take a look at it in Excel. Be sure to delete all prevous _HX4.TXT files before comparing to my output.

[njcurveball]

I am not using Access, but I would like to suggest that you send Ken the extract file. He can run a compare with his export and see if they match.

The piece that I don't see being "apples to apples" is putting the data into Access. If the two files match, than the Access piece is the likely culprit. If they do not, than it is the exporter.

just my thoughts,
Jim

[Rick]

PL5 Exact match

PL4
1 11.875
2 6.3572
3 3.12
4 13.7143
5 50.5013
6 4.15
7 9.3
8 5.0588


That was export PL5 first and then PL4 and open in Excel.

[Rick]

I did it on my other computer and got the same results.

[km]

Thanks Rick, as far as you can tell, is the KLine and K related items the only fields that are not matching the screen? And is PL-4 the only one that has a mis-match?

If the problem is limited to the K items and PL-4, that will help me figure it out.

[MikeDee]

Ken I ran the same test on the same data.

There seem to be some problems with the pace figures. in the PL 4 mode many of the pace fig come up with 0's

I did my test in access but I pasted them to a spreadsheet and attached it
tab 1 is pl5
tab 2 is pl4

like your test I just change the pace line button and re-exported to a clean hx4 txt file and then imported same into access.

[km]

Thanks Mike, obviously that's the problem with the KLine (blanks in those figures). Rick's theory is that there is something amiss when switching to PL-4, not the export. I can't seem to make an error happen though. It should be obvious in this case while looking at the screen, if switching to PL-4, some of the horses will blank out in the PAC-PER columns, but i can't make that happen with race-2 CD Nov12. Maybe it is a specific sequence of clicks or keystrokes of those PLs that sets off the problem :confused:

[MikeDee]

I think I may have found something.

1. all of the PL exports have the problem not just 4 so do 1,2, and 3.

It is some kind of button sequencing problem.

If I just go km screen and click on pl 4 I get the pac numbers that do not match the export.

But when I go to pl 4 and then go to race 1 and then click back to race 2
the pac figures on the screen agree with the pac figs that are exported and they are different then the ones that you see if you just click on the pl button without re-sequencing the race button.

Rick see if you can duplicate this sequencing

Mike

[Rick]

I have noticed that in PL4.

When I use the + key to advance a few races and then use the drop down arrow on the races to go back to race 1. That is where I really notice it.

Will do some more experimenting and will keep track of the keystrokes.

[MikeDee]

if I go to race one and then back to race 2 I get one set of pac figs. ig I go to race 3 and back to race 2 I get a different set of pac figs.

If I go to race 4 -10 and back to race 2 I get the same set of pac figs that I get when I just go to race 2 and toggle the pl button without going to a different race.

[Rick]

It must be the electricity.

3rd computer, same results.

[km]

Aha, i see it on the screen now. If i go to PL-4, then flip back to race-1 and then back to race-2, the PAC-PER figures are blank for some horses in race-2!

Go back to PL-5, and it fixes it.

So it it is the change action between the race and the PL mode i need to look at. Thanks for helping me hunt down this needle in a haystack - i'll figure it out, probably not resetting when there is a multiple actions taking place.

[MikeDee]

That's it Ken and the set of figs that have the blanks are the one that are getting exportedfor pl4

It does not seem to matter what you do on the screens. when you export you get the version with the blanks for pl4.

[km]

I fixed it by resetting all the program variables to zero each time there is a race change or PL change. This is already an aspect to the program, but for some reason, PL-4 was not getting it! I"m having Mike and Rick beta-test now.

I think this bug was recent, probably with the October version after additions to several screens and export were made with the TPG. The TPG requires a complete exit of the standard race and then access to an external file and then return to the race. During that process, there may have been leftover variable assignments from the previous race. So it was very inconsistent and hard to pin down.

Anyway,if it comes up clean, i'll send out a new version to Rick to post if anyone else is interested.

Note that this is not an issue for those of you that export or use the default PL-5 most of the time. Don't bother to download the update. The bug was irregular, but seems to occur if the PL mode is changed at the same time all races on the card are being cycled.

Thanks to Ed (Gambler) for this discovery, and Mike and Rick for patiently tracking it down for me.

[Rick]

And a Turkey Day present for Ed and anyone else that needs it, from km.

This fixes the problem with PL4.

Download PL4 bug fix to HTR2.exe (http://www.homebased2.com/km/dnl/htr2.exe)

Download PL4 bug fix to HTR2006.exe (http://www.homebased2.com/km/dnl/htr2006.exe)

The files are the same, just different names.

Happy Thanksgiving to all.

[BillW]

Thanks for the hard work guys, have a great holiday.

Bill

[Gambler]

I also would like to thank you for all the hard work.