Menu
▾
▴
phpesp-devel
|
From: Roman Y. <ro...@ss...> - 2003-11-03 17:06:08
|
Matthew, I know of Rank type behavior, I'm not talking about that. In my case, I am suing Radio Buttons and only two out of 50+ questions getting messed up! Thanks for your help tho' Roman On Monday, November 3, 2003, at 06:10 PM, Matthew Gregg wrote: > Your problem might be related to this bug: > http://sourceforge.net/tracker/ > index.php?func=detail&aid=827824&group_id=8956&atid=108956 > > Try taking that "off by one" bug into account and see if the > frequencies > work out. > > On Mon, 2003-11-03 at 10:08, Roman Yashin wrote: >> Hi, >> >> >> >> I think there is a possible bug in CSV export function. There is >> always a chance that my dataset is messed up, but I don't think so. >> I'm doing more tests to eliminate that possibility. Meanwhile, has >> anyone experience the following: >> >> >> >> The frequencies for question choices on the report page DO NOT >> correspond to the frequencies for the same question when you analyze >> it in SPSS or Excel? >> >> >> >> Example: >> >> Question: I believe there could a bug in CSV Export function. >> >> >> >> In ESP Report you get: >> >> Strongly Agree 8.8% (13) >> >> Agree 59.9% (88) >> >> Disagree 29.9% (44) >> >> Strongly Disagree 1.4% (2) >> >> TOTAL 100.0% 147 >> >> >> >> In CSV file you get (SPSS and Excel): >> >> Strongly Agree (16) >> >> Agree (78) >> >> Disagree (52) >> >> Strongly Disagree (1) >> >> TOTAL 100.0% 147 >> >> >> >> I should've probably made sure it is a bug before bothering you all, >> but I figured if somebody has experienced something similar it will >> save me a great deal of time. >> >> >> >> Cheers, >> >> >> >> Roman >> >> > -- > mcg > ------------------------------------- > The IT Lab (http://www.itlab.musc.edu) |
|
From: Roman Y. <Ro...@ss...> - 2003-11-04 06:23:29
|
I'm pretty sure now it's a bug. I just imported the same data using ODBC connection and now spreadsheet's numbers correspond to the results on the report page. Maybe it is something in response_select_name()? Any feedback is highly appreciated. Roman -----Original Message----- From: Roman Yashin [mailto:ro...@ss...] Sent: Monday, November 03, 2003 7:06 PM To: Matthew Gregg Cc: php...@li... Subject: Re: [phpesp-dev] CVS export possible bug Matthew, I know of Rank type behavior, I'm not talking about that. In my case, I am suing Radio Buttons and only two out of 50+ questions getting messed up! Thanks for your help tho' Roman On Monday, November 3, 2003, at 06:10 PM, Matthew Gregg wrote: > Your problem might be related to this bug: > http://sourceforge.net/tracker/ > index.php?func=detail&aid=827824&group_id=8956&atid=108956 > > Try taking that "off by one" bug into account and see if the > frequencies > work out. > > On Mon, 2003-11-03 at 10:08, Roman Yashin wrote: >> Hi, >> >> >> >> I think there is a possible bug in CSV export function. There is >> always a chance that my dataset is messed up, but I don't think so. >> I'm doing more tests to eliminate that possibility. Meanwhile, has >> anyone experience the following: >> >> >> >> The frequencies for question choices on the report page DO NOT >> correspond to the frequencies for the same question when you analyze >> it in SPSS or Excel? >> >> >> >> Example: >> >> Question: I believe there could a bug in CSV Export function. >> >> >> >> In ESP Report you get: >> >> Strongly Agree 8.8% (13) >> >> Agree 59.9% (88) >> >> Disagree 29.9% (44) >> >> Strongly Disagree 1.4% (2) >> >> TOTAL 100.0% 147 >> >> >> >> In CSV file you get (SPSS and Excel): >> >> Strongly Agree (16) >> >> Agree (78) >> >> Disagree (52) >> >> Strongly Disagree (1) >> >> TOTAL 100.0% 147 >> >> >> >> I should've probably made sure it is a bug before bothering you all, >> but I figured if somebody has experienced something similar it will >> save me a great deal of time. >> >> >> >> Cheers, >> >> >> >> Roman >> >> > -- > mcg > ------------------------------------- > The IT Lab (http://www.itlab.musc.edu) ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ phpESP-devel mailing list php...@li... https://lists.sourceforge.net/lists/listinfo/phpesp-devel |
|
From: Roman Y. <Ro...@ss...> - 2003-11-06 10:07:00
|
I think I found it! (Haven't fixed it yet tho'). It's in espresponse.inc, response_select_name() function. The problem is in returning $nam array (see line ~651). It's being created couple of lines before with $nam[$key] = $val. I've done the following tests: First: count ($nam,COUNT_RECURSIVE) for $nam[$key] = $val // where $key = question name/label Second: count ($nam,COUNT_RECURSIVE) for $nam[] = $val // notice no variable for index The two arrays have different size, which I believe is incorrect. BTW, the second array has correct data for CSV export. Something happens when we use question names for indexes in the array. I hope all these make sense. I am afraid this behavior might surface somewhere else. James, Kon and others I don't know: What do you guys think about it? Cheers, Roman -----Original Message----- From: Roman Yashin [mailto:ro...@ss...] Sent: Monday, November 03, 2003 7:06 PM To: Matthew Gregg Cc: php...@li... Subject: Re: [phpesp-dev] CVS export possible bug Matthew, I know of Rank type behavior, I'm not talking about that. In my case, I am suing Radio Buttons and only two out of 50+ questions getting messed up! Thanks for your help tho' Roman On Monday, November 3, 2003, at 06:10 PM, Matthew Gregg wrote: > Your problem might be related to this bug: > http://sourceforge.net/tracker/ > index.php?func=detail&aid=827824&group_id=8956&atid=108956 > > Try taking that "off by one" bug into account and see if the > frequencies > work out. > > On Mon, 2003-11-03 at 10:08, Roman Yashin wrote: >> Hi, >> >> >> >> I think there is a possible bug in CSV export function. There is >> always a chance that my dataset is messed up, but I don't think so. >> I'm doing more tests to eliminate that possibility. Meanwhile, has >> anyone experience the following: >> >> >> >> The frequencies for question choices on the report page DO NOT >> correspond to the frequencies for the same question when you analyze >> it in SPSS or Excel? >> >> >> >> Example: >> >> Question: I believe there could a bug in CSV Export function. >> >> >> >> In ESP Report you get: >> >> Strongly Agree 8.8% (13) >> >> Agree 59.9% (88) >> >> Disagree 29.9% (44) >> >> Strongly Disagree 1.4% (2) >> >> TOTAL 100.0% 147 >> >> >> >> In CSV file you get (SPSS and Excel): >> >> Strongly Agree (16) >> >> Agree (78) >> >> Disagree (52) >> >> Strongly Disagree (1) >> >> TOTAL 100.0% 147 >> >> >> >> I should've probably made sure it is a bug before bothering you all, >> but I figured if somebody has experienced something similar it will >> save me a great deal of time. >> >> >> >> Cheers, >> >> >> >> Roman >> >> > -- > mcg > ------------------------------------- > The IT Lab (http://www.itlab.musc.edu) ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ phpESP-devel mailing list php...@li... https://lists.sourceforge.net/lists/listinfo/phpesp-devel |
|
From: Kon <ang...@ip...> - 2003-11-07 10:47:17
|
Hey Roman, I'll have a look at this over the weekend as I'm already doing some redesign work on my devel version. I'll let you know what I find. Regards Kon On Thu, 2003-11-06 at 21:06, Roman Yashin wrote: > I think I found it! (Haven't fixed it yet tho'). It's in > espresponse.inc, response_select_name() function. > > > > The problem is in returning $nam array (see line ~651). It's being > created couple of lines before with $nam[$key] = $val. > > > > I've done the following tests: > > > > First: count ($nam,COUNT_RECURSIVE) for > > $nam[$key] = $val // where $key = question name/label > > > > Second: count ($nam,COUNT_RECURSIVE) for > > $nam[] = $val // notice no variable for index > > > > The two arrays have different size, which I believe is incorrect. BTW, > the second array has correct data for CSV export. Something happens > when we use question names for indexes in the array. > > > > I hope all these make sense. I am afraid this behavior might surface > somewhere else. > > James, Kon and others I don't know: > > What do you guys think about it? > > > > > > Cheers, > > > > Roman > > > > > > > > -----Original Message----- > From: Roman Yashin [mailto:ro...@ss...] > Sent: Monday, November 03, 2003 7:06 PM > To: Matthew Gregg > Cc: php...@li... > Subject: Re: [phpesp-dev] CVS export possible bug > > > > Matthew, > > > > I know of Rank type behavior, I'm not talking about that. > > In my case, I am suing Radio Buttons and only two out of 50+ > questions > > getting messed up! > > > > Thanks for your help tho' > > > > Roman > > > > On Monday, November 3, 2003, at 06:10 PM, Matthew Gregg wrote: > > > > > Your problem might be related to this bug: > > > http://sourceforge.net/tracker/ > > > index.php?func=detail&aid=827824&group_id=8956&atid=108956 > > > > > > Try taking that "off by one" bug into account and see if the > > > frequencies > > > work out. > > > > > > On Mon, 2003-11-03 at 10:08, Roman Yashin wrote: > > >> Hi, > > >> > > >> > > >> > > >> I think there is a possible bug in CSV export function. There is > > >> always a chance that my dataset is messed up, but I don't think so. > > >> I'm doing more tests to eliminate that possibility. Meanwhile, has > > >> anyone experience the following: > > >> > > >> > > >> > > >> The frequencies for question choices on the report page DO NOT > > >> correspond to the frequencies for the same question when you > analyze > > >> it in SPSS or Excel? > > >> > > >> > > >> > > >> Example: > > >> > > >> Question: I believe there could a bug in CSV Export function. > > >> > > >> > > >> > > >> In ESP Report you get: > > >> > > >> Strongly Agree 8.8% (13) > > >> > > >> Agree 59.9% (88) > > >> > > >> Disagree 29.9% (44) > > >> > > >> Strongly Disagree 1.4% (2) > > >> > > >> TOTAL 100.0% 147 > > >> > > >> > > >> > > >> In CSV file you get (SPSS and Excel): > > >> > > >> Strongly Agree (16) > > >> > > >> Agree (78) > > >> > > >> Disagree (52) > > >> > > >> Strongly Disagree (1) > > >> > > >> TOTAL 100.0% 147 > > >> > > >> > > >> > > >> I should've probably made sure it is a bug before bothering you > all, > > >> but I figured if somebody has experienced something similar it will > > >> save me a great deal of time. > > >> > > >> > > >> > > >> Cheers, > > >> > > >> > > >> > > >> Roman > > >> > > >> > > > -- > > > mcg > > > ------------------------------------- > > > The IT Lab (http://www.itlab.musc.edu) > > > > > > > > ------------------------------------------------------- > > This SF.net email is sponsored by: SF.net Giveback Program. > > Does SourceForge.net help you be more productive? Does it > > help you create better code? SHARE THE LOVE, and help us help > > YOU! Click Here: http://sourceforge.net/donate/ > > _______________________________________________ > > phpESP-devel mailing list > > php...@li... > > https://lists.sourceforge.net/lists/listinfo/phpesp-devel > > |
|
From: Roman Y. <Ro...@ss...> - 2003-11-06 14:56:10
|
Update: Had a chance to talk to Matthew (thanks to IRC) and he had a very simple and brilliant idea! I think we've figured out what is happening. When I designed this questionnaire, I let phpESP fill in the question names for me. The problem is that I had some question that were worded similarly, i.e. had different endings with the same beginnings. Don't ask me why, I didn't come up with this questionnaire, but I can image some standardized surveys being like that. This became a problem when the names got trimmed to 30 characters and I tried to export data to CSV. Anyhow, I think this is what happed. I haven't had a chance to look at the part of the code that does the naming trick. I thought I would keep you up-to-date, so you won't be looking in a wrong place. All the best, Roman -----Original Message----- From: Roman Yashin [mailto:Ro...@ss...] Sent: Thursday, November 06, 2003 12:06 PM To: php...@li... Subject: RE: [phpesp-dev] CVS export possible bug I think I found it! (Haven't fixed it yet tho'). It's in espresponse.inc, response_select_name() function. The problem is in returning $nam array (see line ~651). It's being created couple of lines before with $nam[$key] = $val. I've done the following tests: First: count ($nam,COUNT_RECURSIVE) for $nam[$key] = $val // where $key = question name/label Second: count ($nam,COUNT_RECURSIVE) for $nam[] = $val // notice no variable for index The two arrays have different size, which I believe is incorrect. BTW, the second array has correct data for CSV export. Something happens when we use question names for indexes in the array. I hope all these make sense. I am afraid this behavior might surface somewhere else. James, Kon and others I don't know: What do you guys think about it? Cheers, Roman -----Original Message----- From: Roman Yashin [mailto:ro...@ss...] Sent: Monday, November 03, 2003 7:06 PM To: Matthew Gregg Cc: php...@li... Subject: Re: [phpesp-dev] CVS export possible bug Matthew, I know of Rank type behavior, I'm not talking about that. In my case, I am suing Radio Buttons and only two out of 50+ questions getting messed up! Thanks for your help tho' Roman On Monday, November 3, 2003, at 06:10 PM, Matthew Gregg wrote: > Your problem might be related to this bug: > http://sourceforge.net/tracker/ > index.php?func=detail&aid=827824&group_id=8956&atid=108956 > > Try taking that "off by one" bug into account and see if the > frequencies > work out. > > On Mon, 2003-11-03 at 10:08, Roman Yashin wrote: >> Hi, >> >> >> >> I think there is a possible bug in CSV export function. There is >> always a chance that my dataset is messed up, but I don't think so. >> I'm doing more tests to eliminate that possibility. Meanwhile, has >> anyone experience the following: >> >> >> >> The frequencies for question choices on the report page DO NOT >> correspond to the frequencies for the same question when you analyze >> it in SPSS or Excel? >> >> >> >> Example: >> >> Question: I believe there could a bug in CSV Export function. >> >> >> >> In ESP Report you get: >> >> Strongly Agree 8.8% (13) >> >> Agree 59.9% (88) >> >> Disagree 29.9% (44) >> >> Strongly Disagree 1.4% (2) >> >> TOTAL 100.0% 147 >> >> >> >> In CSV file you get (SPSS and Excel): >> >> Strongly Agree (16) >> >> Agree (78) >> >> Disagree (52) >> >> Strongly Disagree (1) >> >> TOTAL 100.0% 147 >> >> >> >> I should've probably made sure it is a bug before bothering you all, >> but I figured if somebody has experienced something similar it will >> save me a great deal of time. >> >> >> >> Cheers, >> >> >> >> Roman >> >> > -- > mcg > ------------------------------------- > The IT Lab (http://www.itlab.musc.edu) ------------------------------------------------------- This SF.net email is sponsored by: SF.net Giveback Program. Does SourceForge.net help you be more productive? Does it help you create better code? SHARE THE LOVE, and help us help YOU! Click Here: http://sourceforge.net/donate/ _______________________________________________ phpESP-devel mailing list php...@li... https://lists.sourceforge.net/lists/listinfo/phpesp-devel |
|
From: Matthew G. <gr...@mu...> - 2003-11-06 15:18:20
|
If this turns out to be the cause, it is already fixed in CVS. See bug #789042 <http://sourceforge.net/tracker/?group_id=8956&atid=108956&func=detail&aid=789042> On Thu, 2003-11-06 at 09:55, Roman Yashin wrote: > Update: > > > > Had a chance to talk to Matthew (thanks to IRC) and he had a very > simple and brilliant idea! I think we've figured out what is > happening. > > When I designed this questionnaire, I let phpESP fill in the question > names for me. The problem is that I had some question that were worded > similarly, i.e. had different endings with the same beginnings. Don't > ask me why, I didn't come up with this questionnaire, but I can image > some standardized surveys being like that. This became a problem when > the names got trimmed to 30 characters and I tried to export data to > CSV. Anyhow, I think this is what happed. I haven't had a chance to > look at the part of the code that does the naming trick. I thought I > would keep you up-to-date, so you won't be looking in a wrong place. > > > > All the best, > > > > Roman > > > > -----Original Message----- > From: Roman Yashin [mailto:Ro...@ss...] > Sent: Thursday, November 06, 2003 12:06 PM > To: php...@li... > Subject: RE: [phpesp-dev] CVS export possible bug > > > > I think I found it! (Haven't fixed it yet tho'). It's in > espresponse.inc, response_select_name() function. > > > > The problem is in returning $nam array (see line ~651). It's being > created couple of lines before with $nam[$key] = $val. > > > > I've done the following tests: > > > > First: count ($nam,COUNT_RECURSIVE) for > > $nam[$key] = $val // where $key = question name/label > > > > Second: count ($nam,COUNT_RECURSIVE) for > > $nam[] = $val // notice no variable for index > > > > The two arrays have different size, which I believe is incorrect. BTW, > the second array has correct data for CSV export. Something happens > when we use question names for indexes in the array. > > > > I hope all these make sense. I am afraid this behavior might surface > somewhere else. > > James, Kon and others I don't know: > > What do you guys think about it? > > > > > > Cheers, > > > > Roman > > > > > > > > -----Original Message----- > From: Roman Yashin [mailto:ro...@ss...] > Sent: Monday, November 03, 2003 7:06 PM > To: Matthew Gregg > Cc: php...@li... > Subject: Re: [phpesp-dev] CVS export possible bug > > > > Matthew, > > > > I know of Rank type behavior, I'm not talking about that. > > In my case, I am suing Radio Buttons and only two out of 50+ > questions > > getting messed up! > > > > Thanks for your help tho' > > > > Roman > > > > On Monday, November 3, 2003, at 06:10 PM, Matthew Gregg wrote: > > > > > Your problem might be related to this bug: > > > http://sourceforge.net/tracker/ > > > index.php?func=detail&aid=827824&group_id=8956&atid=108956 > > > > > > Try taking that "off by one" bug into account and see if the > > > frequencies > > > work out. > > > > > > On Mon, 2003-11-03 at 10:08, Roman Yashin wrote: > > >> Hi, > > >> > > >> > > >> > > >> I think there is a possible bug in CSV export function. There is > > >> always a chance that my dataset is messed up, but I don't think so. > > >> I'm doing more tests to eliminate that possibility. Meanwhile, has > > >> anyone experience the following: > > >> > > >> > > >> > > >> The frequencies for question choices on the report page DO NOT > > >> correspond to the frequencies for the same question when you > analyze > > >> it in SPSS or Excel? > > >> > > >> > > >> > > >> Example: > > >> > > >> Question: I believe there could a bug in CSV Export function. > > >> > > >> > > >> > > >> In ESP Report you get: > > >> > > >> Strongly Agree 8.8% (13) > > >> > > >> Agree 59.9% (88) > > >> > > >> Disagree 29.9% (44) > > >> > > >> Strongly Disagree 1.4% (2) > > >> > > >> TOTAL 100.0% 147 > > >> > > >> > > >> > > >> In CSV file you get (SPSS and Excel): > > >> > > >> Strongly Agree (16) > > >> > > >> Agree (78) > > >> > > >> Disagree (52) > > >> > > >> Strongly Disagree (1) > > >> > > >> TOTAL 100.0% 147 > > >> > > >> > > >> > > >> I should've probably made sure it is a bug before bothering you > all, > > >> but I figured if somebody has experienced something similar it will > > >> save me a great deal of time. > > >> > > >> > > >> > > >> Cheers, > > >> > > >> > > >> > > >> Roman > > >> > > >> > > > -- > > > mcg > > > ------------------------------------- > > > The IT Lab (http://www.itlab.musc.edu) > > > > > > > > ------------------------------------------------------- > > This SF.net email is sponsored by: SF.net Giveback Program. > > Does SourceForge.net help you be more productive? Does it > > help you create better code? SHARE THE LOVE, and help us help > > YOU! Click Here: http://sourceforge.net/donate/ > > _______________________________________________ > > phpESP-devel mailing list > > php...@li... > > https://lists.sourceforge.net/lists/listinfo/phpesp-devel -- mcg ------------------------------------- The IT Lab (http://www.itlab.musc.edu) |
×
Want the latest updates on software, tech news, and AI?
Get latest updates about software, tech news, and AI from SourceForge directly in your inbox once a month.
Thanks for helping keep SourceForge clean.
X