Discussion: View Thread

I'm providing this for  humor and fond (for some of us) memories of how we graphed prior to R (and prior to S and S-PLUS). Germane to the humor as I've announced , I'm retiring. And recently I was organizing my storage and archives. And  I made a modest discovery. (See Attachments).

I discovered the original acetates for overheads (prior to the existence of PowerPoint) for what I'm sure is the very first scatterplot graph that I ever prepared on a computer and printed on a printer. The alternative was purchasing lined graph paper at the stationary store and plotting by hand in pencil. In that decade, at our large Pharma corporation, we used a DEC-10 computer ( we narrowly escaped using an IBM 360). The PC of the day might have been the Osborne, and so-called portable because who doesn't think a several thousand dollar $$$, and about 25 pound or so, portable computer is - portable? More likely than not, one of several SAS programmers in our department was working with me and she actually had figured out how to print a graph from SAS with color and  symbols. Legend and folklore that I can't confirm,  have it that the SAS source code for GPLOT was the original code written by Jim Goodnight (or perhaps John Sall) . More importantly, A SAS programmer who could actually draw symbols and a legend on a graph in color were highly sought after, The alternatives to SAS were BMDP and SPSS.

An oversimplified explanation of the graph. This was a plot of eight dose (x axis) groups of a drug substance (Mycophenolic acid MPA) that had immunosuppressive properties from what we called phase I . (and based on the results from this study we skipped phase II entirely and started the Phase III). Plotted is the pK data (AUC's and cmax) obtained from the first 50 or so patients "first in human" FIH, to ever receive the drug. The plot is intended to show that as dose increases, the number of immunologic organ "rejections" ( T cell mediated) declines. The drug itself was the ester of MPA with the sonorous name of "Mycophenolate Mofetil". The data from this study was used to select the doses for phase III (in the label 2 mg and 3 mg ) and the name MPA was changed to CELLCEPT for solid organ (kidney heart or liver) transplant. (one organ at a time!) .  Back in the day, graphs were Easy-peasy! A colleague thought the marketing people might like to see the graph and the marketing people based on the graph relayed a message to me, my manager, our CEO and other executives and board of directors (definitely not in that order)  that the drug was surely going to be a "Billion-dollar drug". The drug was ultimately approved a few years after this graph was created, and I'm told was very successful "first to market" and I'm told quite profitable

I am reasonably certain this was a "proc plot" and that Gplot did not exist or was in a beta testing mode. There appears to be some jittering. A notable flaw of the graph is the absence of a "source line." Graphs in a submission to FDA , if any, tended to be few and far between, given the expense of printing and plotting.  And many at the corporation held the view that Proc plot type output was not sufficiently accurate for sending to FDA. Someone apparently compared the original raw data to the points on the plot and decided SAS display was not 'sufficiently accurate" and thus  proc plot graph output was never  to be sent  to FDA. There were no PC's and no inkjet printers. In that era one could purchase an HP (Hewlett Packard) large flatbed pen plotter. with pens at the end of the arm (similar to the arm on a first-generation disk drive), - and draw the picture. Those plotters  were considered prohibitively expensive plotters and may have required their own specific plotting language to make the plot. Of course, now, in R, one or two lines of R code , and many switches in the function call and voila the plot with an access to a full palette of colors and output to multiple graph formats (jpg etc.) is prepared in minutes and may in principle be sent to FDA.

Ultimately The graph reminds me  of a variation on a  punchline to a joke "We had it tough. back in the day we used to walk to school, in the winter, barefoot and backwards." 😊

------------------------------
Chris Barker, Ph.D.
2023 Chair Statistical Consulting Section
Consultant and
Adjunct Associate Professor of Biostatistics
www.barkerstats.com


---
"In composition you have all the time you want to decide what to say in 15 seconds, in improvisation you have 15 seconds."
-Steve Lacy
------------------------------

Chris,

Thanks for the humor and history lesson!  1995 was during my undergrad years, and your post reminds of how plots we created were relatively primitive compared to what we can do now.  Your two PDFs look great relative to what I recall making in class, and your anecdote provides a good perspective on what was involved.  Your mention of acetate overheads brings back many memories of creating slides for classroom presentations.

History is great for providing perspective on what you currently possess -- and, in the process, generating gratitude for the present.  It's one of the reasons why I like Stephen Stigler's books so much.  Your post definitely brings back many memories of how it was done and how far graphing has come. 

Thanks again!

I'm providing this for  humor and fond (for some of us) memories of how we graphed prior to R (and prior to S and S-PLUS). Germane to the humor as I've announced , I'm retiring. And recently I was organizing my storage and archives. And  I made a modest discovery. (See Attachments).

I discovered the original acetates for overheads (prior to the existence of PowerPoint) for what I'm sure is the very first scatterplot graph that I ever prepared on a computer and printed on a printer. The alternative was purchasing lined graph paper at the stationary store and plotting by hand in pencil. In that decade, at our large Pharma corporation, we used a DEC-10 computer ( we narrowly escaped using an IBM 360). The PC of the day might have been the Osborne, and so-called portable because who doesn't think a several thousand dollar $$$, and about 25 pound or so, portable computer is - portable? More likely than not, one of several SAS programmers in our department was working with me and she actually had figured out how to print a graph from SAS with color and  symbols. Legend and folklore that I can't confirm,  have it that the SAS source code for GPLOT was the original code written by Jim Goodnight (or perhaps John Sall) . More importantly, A SAS programmer who could actually draw symbols and a legend on a graph in color were highly sought after, The alternatives to SAS were BMDP and SPSS.

An oversimplified explanation of the graph. This was a plot of eight dose (x axis) groups of a drug substance (Mycophenolic acid MPA) that had immunosuppressive properties from what we called phase I . (and based on the results from this study we skipped phase II entirely and started the Phase III). Plotted is the pK data (AUC's and cmax) obtained from the first 50 or so patients "first in human" FIH, to ever receive the drug. The plot is intended to show that as dose increases, the number of immunologic organ "rejections" ( T cell mediated) declines. The drug itself was the ester of MPA with the sonorous name of "Mycophenolate Mofetil". The data from this study was used to select the doses for phase III (in the label 2 mg and 3 mg ) and the name MPA was changed to CELLCEPT for solid organ (kidney heart or liver) transplant. (one organ at a time!) .  Back in the day, graphs were Easy-peasy! A colleague thought the marketing people might like to see the graph and the marketing people based on the graph relayed a message to me, my manager, our CEO and other executives and board of directors (definitely not in that order)  that the drug was surely going to be a "Billion-dollar drug". The drug was ultimately approved a few years after this graph was created, and I'm told was very successful "first to market" and I'm told quite profitable

I am reasonably certain this was a "proc plot" and that Gplot did not exist or was in a beta testing mode. There appears to be some jittering. A notable flaw of the graph is the absence of a "source line." Graphs in a submission to FDA , if any, tended to be few and far between, given the expense of printing and plotting.  And many at the corporation held the view that Proc plot type output was not sufficiently accurate for sending to FDA. Someone apparently compared the original raw data to the points on the plot and decided SAS display was not 'sufficiently accurate" and thus  proc plot graph output was never  to be sent  to FDA. There were no PC's and no inkjet printers. In that era one could purchase an HP (Hewlett Packard) large flatbed pen plotter. with pens at the end of the arm (similar to the arm on a first-generation disk drive), - and draw the picture. Those plotters  were considered prohibitively expensive plotters and may have required their own specific plotting language to make the plot. Of course, now, in R, one or two lines of R code , and many switches in the function call and voila the plot with an access to a full palette of colors and output to multiple graph formats (jpg etc.) is prepared in minutes and may in principle be sent to FDA.

Ultimately The graph reminds me  of a variation on a  punchline to a joke "We had it tough. back in the day we used to walk to school, in the winter, barefoot and backwards." 😊

------------------------------
Chris Barker, Ph.D.
2023 Chair Statistical Consulting Section
Consultant and
Adjunct Associate Professor of Biostatistics
www.barkerstats.com


---
"In composition you have all the time you want to decide what to say in 15 seconds, in improvisation you have 15 seconds."
-Steve Lacy
------------------------------

Chris,

Thanks for the humor and history lesson!  1995 was during my undergrad years, and your post reminds of how plots we created were relatively primitive compared to what we can do now.  Your two PDFs look great relative to what I recall making in class, and your anecdote provides a good perspective on what was involved.  Your mention of acetate overheads brings back many memories of creating slides for classroom presentations.

History is great for providing perspective on what you currently possess -- and, in the process, generating gratitude for the present.  It's one of the reasons why I like Stephen Stigler's books so much.  Your post definitely brings back many memories of how it was done and how far graphing has come. 

Thanks again!

I'm providing this for  humor and fond (for some of us) memories of how we graphed prior to R (and prior to S and S-PLUS). Germane to the humor as I've announced , I'm retiring. And recently I was organizing my storage and archives. And  I made a modest discovery. (See Attachments).

I discovered the original acetates for overheads (prior to the existence of PowerPoint) for what I'm sure is the very first scatterplot graph that I ever prepared on a computer and printed on a printer. The alternative was purchasing lined graph paper at the stationary store and plotting by hand in pencil. In that decade, at our large Pharma corporation, we used a DEC-10 computer ( we narrowly escaped using an IBM 360). The PC of the day might have been the Osborne, and so-called portable because who doesn't think a several thousand dollar $$$, and about 25 pound or so, portable computer is - portable? More likely than not, one of several SAS programmers in our department was working with me and she actually had figured out how to print a graph from SAS with color and  symbols. Legend and folklore that I can't confirm,  have it that the SAS source code for GPLOT was the original code written by Jim Goodnight (or perhaps John Sall) . More importantly, A SAS programmer who could actually draw symbols and a legend on a graph in color were highly sought after, The alternatives to SAS were BMDP and SPSS.

An oversimplified explanation of the graph. This was a plot of eight dose (x axis) groups of a drug substance (Mycophenolic acid MPA) that had immunosuppressive properties from what we called phase I . (and based on the results from this study we skipped phase II entirely and started the Phase III). Plotted is the pK data (AUC's and cmax) obtained from the first 50 or so patients "first in human" FIH, to ever receive the drug. The plot is intended to show that as dose increases, the number of immunologic organ "rejections" ( T cell mediated) declines. The drug itself was the ester of MPA with the sonorous name of "Mycophenolate Mofetil". The data from this study was used to select the doses for phase III (in the label 2 mg and 3 mg ) and the name MPA was changed to CELLCEPT for solid organ (kidney heart or liver) transplant. (one organ at a time!) .  Back in the day, graphs were Easy-peasy! A colleague thought the marketing people might like to see the graph and the marketing people based on the graph relayed a message to me, my manager, our CEO and other executives and board of directors (definitely not in that order)  that the drug was surely going to be a "Billion-dollar drug". The drug was ultimately approved a few years after this graph was created, and I'm told was very successful "first to market" and I'm told quite profitable

I am reasonably certain this was a "proc plot" and that Gplot did not exist or was in a beta testing mode. There appears to be some jittering. A notable flaw of the graph is the absence of a "source line." Graphs in a submission to FDA , if any, tended to be few and far between, given the expense of printing and plotting.  And many at the corporation held the view that Proc plot type output was not sufficiently accurate for sending to FDA. Someone apparently compared the original raw data to the points on the plot and decided SAS display was not 'sufficiently accurate" and thus  proc plot graph output was never  to be sent  to FDA. There were no PC's and no inkjet printers. In that era one could purchase an HP (Hewlett Packard) large flatbed pen plotter. with pens at the end of the arm (similar to the arm on a first-generation disk drive), - and draw the picture. Those plotters  were considered prohibitively expensive plotters and may have required their own specific plotting language to make the plot. Of course, now, in R, one or two lines of R code , and many switches in the function call and voila the plot with an access to a full palette of colors and output to multiple graph formats (jpg etc.) is prepared in minutes and may in principle be sent to FDA.

Ultimately The graph reminds me  of a variation on a  punchline to a joke "We had it tough. back in the day we used to walk to school, in the winter, barefoot and backwards." 😊

------------------------------
Chris Barker, Ph.D.
2023 Chair Statistical Consulting Section
Consultant and
Adjunct Associate Professor of Biostatistics
www.barkerstats.com


---
"In composition you have all the time you want to decide what to say in 15 seconds, in improvisation you have 15 seconds."
-Steve Lacy
------------------------------

Chris,

Thanks for the humor and history lesson!  1995 was during my undergrad years, and your post reminds of how plots we created were relatively primitive compared to what we can do now.  Your two PDFs look great relative to what I recall making in class, and your anecdote provides a good perspective on what was involved.  Your mention of acetate overheads brings back many memories of creating slides for classroom presentations.

History is great for providing perspective on what you currently possess -- and, in the process, generating gratitude for the present.  It's one of the reasons why I like Stephen Stigler's books so much.  Your post definitely brings back many memories of how it was done and how far graphing has come. 

Thanks again!

I'm providing this for  humor and fond (for some of us) memories of how we graphed prior to R (and prior to S and S-PLUS). Germane to the humor as I've announced , I'm retiring. And recently I was organizing my storage and archives. And  I made a modest discovery. (See Attachments).

I discovered the original acetates for overheads (prior to the existence of PowerPoint) for what I'm sure is the very first scatterplot graph that I ever prepared on a computer and printed on a printer. The alternative was purchasing lined graph paper at the stationary store and plotting by hand in pencil. In that decade, at our large Pharma corporation, we used a DEC-10 computer ( we narrowly escaped using an IBM 360). The PC of the day might have been the Osborne, and so-called portable because who doesn't think a several thousand dollar $$$, and about 25 pound or so, portable computer is - portable? More likely than not, one of several SAS programmers in our department was working with me and she actually had figured out how to print a graph from SAS with color and  symbols. Legend and folklore that I can't confirm,  have it that the SAS source code for GPLOT was the original code written by Jim Goodnight (or perhaps John Sall) . More importantly, A SAS programmer who could actually draw symbols and a legend on a graph in color were highly sought after, The alternatives to SAS were BMDP and SPSS.

An oversimplified explanation of the graph. This was a plot of eight dose (x axis) groups of a drug substance (Mycophenolic acid MPA) that had immunosuppressive properties from what we called phase I . (and based on the results from this study we skipped phase II entirely and started the Phase III). Plotted is the pK data (AUC's and cmax) obtained from the first 50 or so patients "first in human" FIH, to ever receive the drug. The plot is intended to show that as dose increases, the number of immunologic organ "rejections" ( T cell mediated) declines. The drug itself was the ester of MPA with the sonorous name of "Mycophenolate Mofetil". The data from this study was used to select the doses for phase III (in the label 2 mg and 3 mg ) and the name MPA was changed to CELLCEPT for solid organ (kidney heart or liver) transplant. (one organ at a time!) .  Back in the day, graphs were Easy-peasy! A colleague thought the marketing people might like to see the graph and the marketing people based on the graph relayed a message to me, my manager, our CEO and other executives and board of directors (definitely not in that order)  that the drug was surely going to be a "Billion-dollar drug". The drug was ultimately approved a few years after this graph was created, and I'm told was very successful "first to market" and I'm told quite profitable

I am reasonably certain this was a "proc plot" and that Gplot did not exist or was in a beta testing mode. There appears to be some jittering. A notable flaw of the graph is the absence of a "source line." Graphs in a submission to FDA , if any, tended to be few and far between, given the expense of printing and plotting.  And many at the corporation held the view that Proc plot type output was not sufficiently accurate for sending to FDA. Someone apparently compared the original raw data to the points on the plot and decided SAS display was not 'sufficiently accurate" and thus  proc plot graph output was never  to be sent  to FDA. There were no PC's and no inkjet printers. In that era one could purchase an HP (Hewlett Packard) large flatbed pen plotter. with pens at the end of the arm (similar to the arm on a first-generation disk drive), - and draw the picture. Those plotters  were considered prohibitively expensive plotters and may have required their own specific plotting language to make the plot. Of course, now, in R, one or two lines of R code , and many switches in the function call and voila the plot with an access to a full palette of colors and output to multiple graph formats (jpg etc.) is prepared in minutes and may in principle be sent to FDA.

Ultimately The graph reminds me  of a variation on a  punchline to a joke "We had it tough. back in the day we used to walk to school, in the winter, barefoot and backwards." 😊

------------------------------
Chris Barker, Ph.D.
2023 Chair Statistical Consulting Section
Consultant and
Adjunct Associate Professor of Biostatistics
www.barkerstats.com


---
"In composition you have all the time you want to decide what to say in 15 seconds, in improvisation you have 15 seconds."
-Steve Lacy
------------------------------