Problems with SPSS Charts
Locked
 
|
Hi all,
We just upgraded from SPSS 13 to 16 and have experienced problems with generating charts of any type. Our database consists of 6.500.000 clients whom we have separated in two .sav files, postpaid customers who sum about 1.750.000 and prepaid customers who add up to about 4.750.000 clients respectively. For example, I would like to create a boxplot with each of the .sav files. In the case of the postpaid file SPSS generates the chart without problems, but after editing the chart I receive the message "Image Rendering Error". With the prepaid file SPSS isn't able to create the chart from the start and delivers the same message "Image Rendering Error" immediately. Besides this error, I haven't found a way to add the mean to a boxplot. Can someone help with these two issues, the image rendering error and adding the mean to a boxplot? Thanks in advance, Vito Palligas Dirección de Marketing Telefónica Móviles de Chile Fijo: +56 2 339 5412 Móvil: +56 9 9225 0491 ====================To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
Anyone have any suggestions?
________________________________ De: Vito Palligas Enviado el: Martes, 03 de Junio de 2008 12:25 Para: '[hidden email]' CC: Carlos Miguel Ide Aravena Asunto: Problems with SPSS Charts Hi all, We just upgraded from SPSS 13 to 16 and have experienced problems with generating charts of any type. Our database consists of 6.500.000 clients whom we have separated in two .sav files, postpaid customers who sum about 1.750.000 and prepaid customers who add up to about 4.750.000 clients respectively. For example, I would like to create a boxplot with each of the .sav files. In the case of the postpaid file SPSS generates the chart without problems, but after editing the chart I receive the message "Image Rendering Error". With the prepaid file SPSS isn't able to create the chart from the start and delivers the same message "Image Rendering Error" immediately. Besides this error, I haven't found a way to add the mean to a boxplot. Can someone help with these two issues, the image rendering error and adding the mean to a boxplot? Thanks in advance, Vito Palligas Dirección de Marketing Telefónica Móviles de Chile Fijo: +56 2 339 5412 Móvil: +56 9 9225 0491 ====================To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
Boxplot requires that all of the cases be stored within the plot and that
might be causing the problems since the technology used for the plot has changed. My own approach would be to sample cases down to a something like 10,000 observations. Try this and run EXPLORE to find the hinges for sample and population. -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Vito Palligas Sent: Wednesday, June 04, 2008 7:52 AM To: [hidden email] Subject: Re: Problems with SPSS Charts Anyone have any suggestions? ________________________________ De: Vito Palligas Enviado el: Martes, 03 de Junio de 2008 12:25 Para: '[hidden email]' CC: Carlos Miguel Ide Aravena Asunto: Problems with SPSS Charts Hi all, We just upgraded from SPSS 13 to 16 and have experienced problems with generating charts of any type. Our database consists of 6.500.000 clients whom we have separated in two .sav files, postpaid customers who sum about 1.750.000 and prepaid customers who add up to about 4.750.000 clients respectively. For example, I would like to create a boxplot with each of the .sav files. In the case of the postpaid file SPSS generates the chart without problems, but after editing the chart I receive the message "Image Rendering Error". With the prepaid file SPSS isn't able to create the chart from the start and delivers the same message "Image Rendering Error" immediately. Besides this error, I haven't found a way to add the mean to a boxplot. Can someone help with these two issues, the image rendering error and adding the mean to a boxplot? Thanks in advance, Vito Palligas Dirección de Marketing Telefónica Móviles de Chile Fijo: +56 2 339 5412 Móvil: +56 9 9225 0491 ======= To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
Thanks for your response. Referring to the other topic, do you know of a way to add the mean to the boxplot? I've reviewed the options and haven't found a way to do this using the dialogue boxes. Perhaps something like this can be done using the GPL interface.
Regards, Vito -----Mensaje original----- De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de ViAnn Beadle Enviado el: Miércoles, 04 de Junio de 2008 10:55 Para: [hidden email] Asunto: Re: Problems with SPSS Charts Boxplot requires that all of the cases be stored within the plot and that might be causing the problems since the technology used for the plot has changed. My own approach would be to sample cases down to a something like 10,000 observations. Try this and run EXPLORE to find the hinges for sample and population. -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Vito Palligas Sent: Wednesday, June 04, 2008 7:52 AM To: [hidden email] Subject: Re: Problems with SPSS Charts Anyone have any suggestions? ________________________________ De: Vito Palligas Enviado el: Martes, 03 de Junio de 2008 12:25 Para: '[hidden email]' CC: Carlos Miguel Ide Aravena Asunto: Problems with SPSS Charts Hi all, We just upgraded from SPSS 13 to 16 and have experienced problems with generating charts of any type. Our database consists of 6.500.000 clients whom we have separated in two .sav files, postpaid customers who sum about 1.750.000 and prepaid customers who add up to about 4.750.000 clients respectively. For example, I would like to create a boxplot with each of the .sav files. In the case of the postpaid file SPSS generates the chart without problems, but after editing the chart I receive the message "Image Rendering Error". With the prepaid file SPSS isn't able to create the chart from the start and delivers the same message "Image Rendering Error" immediately. Besides this error, I haven't found a way to add the mean to a boxplot. Can someone help with these two issues, the image rendering error and adding the mean to a boxplot? Thanks in advance, Vito Palligas Dirección de Marketing Telefónica Móviles de Chile Fijo: +56 2 339 5412 Móvil: +56 9 9225 0491 ======= To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
You can add the mean to a boxplot with GPL using an extra ELEMENT statement and perhaps a secondary dataset, but the logic of a boxplot is all based on order statistics. A mean line could be rather misleading, which is why there is no built-in way to do it. Are you sure you want to do this?
Regards, Jon Peck -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Vito Palligas Sent: Wednesday, June 04, 2008 9:46 AM To: [hidden email] Subject: Re: [SPSSX-L] Problems with SPSS Charts Thanks for your response. Referring to the other topic, do you know of a way to add the mean to the boxplot? I've reviewed the options and haven't found a way to do this using the dialogue boxes. Perhaps something like this can be done using the GPL interface. Regards, Vito -----Mensaje original----- De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de ViAnn Beadle Enviado el: Miércoles, 04 de Junio de 2008 10:55 Para: [hidden email] Asunto: Re: Problems with SPSS Charts Boxplot requires that all of the cases be stored within the plot and that might be causing the problems since the technology used for the plot has changed. My own approach would be to sample cases down to a something like 10,000 observations. Try this and run EXPLORE to find the hinges for sample and population. -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Vito Palligas Sent: Wednesday, June 04, 2008 7:52 AM To: [hidden email] Subject: Re: Problems with SPSS Charts Anyone have any suggestions? ________________________________ De: Vito Palligas Enviado el: Martes, 03 de Junio de 2008 12:25 Para: '[hidden email]' CC: Carlos Miguel Ide Aravena Asunto: Problems with SPSS Charts Hi all, We just upgraded from SPSS 13 to 16 and have experienced problems with generating charts of any type. Our database consists of 6.500.000 clients whom we have separated in two .sav files, postpaid customers who sum about 1.750.000 and prepaid customers who add up to about 4.750.000 clients respectively. For example, I would like to create a boxplot with each of the .sav files. In the case of the postpaid file SPSS generates the chart without problems, but after editing the chart I receive the message "Image Rendering Error". With the prepaid file SPSS isn't able to create the chart from the start and delivers the same message "Image Rendering Error" immediately. Besides this error, I haven't found a way to add the mean to a boxplot. Can someone help with these two issues, the image rendering error and adding the mean to a boxplot? Thanks in advance, Vito Palligas Dirección de Marketing Telefónica Móviles de Chile Fijo: +56 2 339 5412 Móvil: +56 9 9225 0491 ======= To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
Thanks for the guidance. I understand why this could be misleading. This format of boxplot is part of a report with the company's monthly indicators. The chart has been created in the past versions of the report using Minitab which has a built in way to this and the idea is to migrate all the processes to SPSS.
The main purpose behind wanting to do this is to see how far the median and the mean are apart as a visual indicator of skewness in the distribution of the different indicators presented in the report. I am just recently becoming acquainted with SPSS 16 and since we upgraded from version 13, I haven't had the chance to learn GPL. I am quite comfortable with SPSS syntax, so how difficult would you say it is to develop this kind of plot for an intermediate syntax user like me? Regards, Vito -----Mensaje original----- De: Peck, Jon [mailto:[hidden email]] Enviado el: Miércoles, 04 de Junio de 2008 12:07 Para: Vito Palligas; [hidden email] Asunto: RE: Re: [SPSSX-L] Problems with SPSS Charts You can add the mean to a boxplot with GPL using an extra ELEMENT statement and perhaps a secondary dataset, but the logic of a boxplot is all based on order statistics. A mean line could be rather misleading, which is why there is no built-in way to do it. Are you sure you want to do this? Regards, Jon Peck -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Vito Palligas Sent: Wednesday, June 04, 2008 9:46 AM To: [hidden email] Subject: Re: [SPSSX-L] Problems with SPSS Charts Thanks for your response. Referring to the other topic, do you know of a way to add the mean to the boxplot? I've reviewed the options and haven't found a way to do this using the dialogue boxes. Perhaps something like this can be done using the GPL interface. Regards, Vito -----Mensaje original----- De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de ViAnn Beadle Enviado el: Miércoles, 04 de Junio de 2008 10:55 Para: [hidden email] Asunto: Re: Problems with SPSS Charts Boxplot requires that all of the cases be stored within the plot and that might be causing the problems since the technology used for the plot has changed. My own approach would be to sample cases down to a something like 10,000 observations. Try this and run EXPLORE to find the hinges for sample and population. -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Vito Palligas Sent: Wednesday, June 04, 2008 7:52 AM To: [hidden email] Subject: Re: Problems with SPSS Charts Anyone have any suggestions? ________________________________ De: Vito Palligas Enviado el: Martes, 03 de Junio de 2008 12:25 Para: '[hidden email]' CC: Carlos Miguel Ide Aravena Asunto: Problems with SPSS Charts Hi all, We just upgraded from SPSS 13 to 16 and have experienced problems with generating charts of any type. Our database consists of 6.500.000 clients whom we have separated in two .sav files, postpaid customers who sum about 1.750.000 and prepaid customers who add up to about 4.750.000 clients respectively. For example, I would like to create a boxplot with each of the .sav files. In the case of the postpaid file SPSS generates the chart without problems, but after editing the chart I receive the message "Image Rendering Error". With the prepaid file SPSS isn't able to create the chart from the start and delivers the same message "Image Rendering Error" immediately. Besides this error, I haven't found a way to add the mean to a boxplot. Can someone help with these two issues, the image rendering error and adding the mean to a boxplot? Thanks in advance, Vito Palligas Dirección de Marketing Telefónica Móviles de Chile Fijo: +56 2 339 5412 Móvil: +56 9 9225 0491 ======= To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
|
|
I happen to have a syntax file of variations on boxplots with means that I will send you offline. It uses the cars.sav dataset shipped with SPSS.
-----Original Message----- From: Vito Palligas [mailto:[hidden email]] Sent: Wednesday, June 04, 2008 11:21 AM To: Peck, Jon; [hidden email] Cc: Carlos Miguel Ide Aravena Subject: RE: Re: [SPSSX-L] Problems with SPSS Charts Thanks for the guidance. I understand why this could be misleading. This format of boxplot is part of a report with the company's monthly indicators. The chart has been created in the past versions of the report using Minitab which has a built in way to this and the idea is to migrate all the processes to SPSS. The main purpose behind wanting to do this is to see how far the median and the mean are apart as a visual indicator of skewness in the distribution of the different indicators presented in the report. I am just recently becoming acquainted with SPSS 16 and since we upgraded from version 13, I haven't had the chance to learn GPL. I am quite comfortable with SPSS syntax, so how difficult would you say it is to develop this kind of plot for an intermediate syntax user like me? Regards, Vito ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
AW: Problems with SPSS Charts
|
|
In reply to this post by Vito Palligas
Hi Vito
here are several examples for box plots I have save from somewhere I don't remember. Sorry, but I can't remember the author of the code. Anyway, al three examples are box plots with mean markers. They all use the data file cars.sav. that comes with SPSS. Hope this helps, Christian *The chart might be easier to read with a point marker instead of a line: * boxplot with mean marker. GGRAPH /GRAPHDATASET NAME="graphdataset" VARIABLES=origin horse /GRAPHSPEC SOURCE=INLINE. BEGIN GPL SOURCE: s=userSource(id("graphdataset")) DATA: origin=col(source(s), name("origin"), unit.category()) DATA: horse=col(source(s), name("horse")) GUIDE: axis(dim(1), label("Country of Origin")) GUIDE: axis(dim(2), label("Horsepower")) SCALE: cat(dim(1), include("1", "2", "3")) SCALE: linear(dim(2), include(0)) ELEMENT: schema(position(bin.quantile.letter(origin*horse)),size(size."30px")) ELEMENT: point(position(summary.mean(origin*horse)), shape(shape.square)) END GPL. * boxplot with mean marker. GGRAPH /GRAPHDATASET NAME="graphdataset" VARIABLES=origin horse /GRAPHSPEC SOURCE=INLINE. BEGIN GPL SOURCE: s=userSource(id("graphdataset")) DATA: origin=col(source(s), name("origin"), unit.category()) DATA: horse=col(source(s), name("horse")) GUIDE: axis(dim(1), label("Country of Origin")) GUIDE: axis(dim(2), label("Horsepower")) SCALE: cat(dim(1), include("1", "2", "3")) SCALE: linear(dim(2), include(0)) ELEMENT: schema(position(bin.quantile.letter(origin*horse)),size(size."30px")) ELEMENT: interval(position(region.spread.range(summary.mean(origin*horse))),size(size ."20px")) END GPL. * You could, instead, add a line connecting the means across the boxes, but this * would only make sense if the categories are ordered. Here is a version with * marker and connecting line. * boxplot with mean marker. GGRAPH /GRAPHDATASET NAME="graphdataset" VARIABLES=origin horse /GRAPHSPEC SOURCE=INLINE. BEGIN GPL SOURCE: s=userSource(id("graphdataset")) DATA: origin=col(source(s), name("origin"), unit.category()) DATA: horse=col(source(s), name("horse")) GUIDE: axis(dim(1), label("Country of Origin")) GUIDE: axis(dim(2), label("Horsepower")) SCALE: cat(dim(1), include("1", "2", "3")) SCALE: linear(dim(2), include(0)) ELEMENT: schema(position(bin.quantile.letter(origin*horse))) ELEMENT: point(position(summary.mean(origin*horse)),shape(shape.square)) ELEMENT: interval(position(region.spread.range(summary.mean(origin*horse)))) END GPL. -----Ursprüngliche Nachricht----- Von: SPSSX(r) Discussion [mailto:[hidden email]]Im Auftrag von Vito Palligas Gesendet: Mittwoch, 4. Juni 2008 17:46 An: [hidden email] Betreff: Re: Problems with SPSS Charts Thanks for your response. Referring to the other topic, do you know of a way to add the mean to the boxplot? I've reviewed the options and haven't found a way to do this using the dialogue boxes. Perhaps something like this can be done using the GPL interface. Regards, Vito -----Mensaje original----- De: SPSSX(r) Discussion [mailto:[hidden email]] En nombre de ViAnn Beadle Enviado el: Miércoles, 04 de Junio de 2008 10:55 Para: [hidden email] Asunto: Re: Problems with SPSS Charts Boxplot requires that all of the cases be stored within the plot and that might be causing the problems since the technology used for the plot has changed. My own approach would be to sample cases down to a something like 10,000 observations. Try this and run EXPLORE to find the hinges for sample and population. -----Original Message----- From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Vito Palligas Sent: Wednesday, June 04, 2008 7:52 AM To: [hidden email] Subject: Re: Problems with SPSS Charts Anyone have any suggestions? ________________________________ De: Vito Palligas Enviado el: Martes, 03 de Junio de 2008 12:25 Para: '[hidden email]' CC: Carlos Miguel Ide Aravena Asunto: Problems with SPSS Charts Hi all, We just upgraded from SPSS 13 to 16 and have experienced problems with generating charts of any type. Our database consists of 6.500.000 clients whom we have separated in two .sav files, postpaid customers who sum about 1.750.000 and prepaid customers who add up to about 4.750.000 clients respectively. For example, I would like to create a boxplot with each of the .sav files. In the case of the postpaid file SPSS generates the chart without problems, but after editing the chart I receive the message "Image Rendering Error". With the prepaid file SPSS isn't able to create the chart from the start and delivers the same message "Image Rendering Error" immediately. Besides this error, I haven't found a way to add the mean to a boxplot. Can someone help with these two issues, the image rendering error and adding the mean to a boxplot? Thanks in advance, Vito Palligas Dirección de Marketing Telefónica Móviles de Chile Fijo: +56 2 339 5412 Móvil: +56 9 9225 0491 ======= To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD |
«
Return to SPSSX Discussion
|
1 view|%1 views
| Free forum by Nabble | Edit this page |