A Macro Automates your Table 1 Geliang Gan Biostatistician YCAS - PowerPoint PPT Presentation

Yale Center for Analytical Sciences (YCAS) A Macro Automates your Table 1 Geliang Gan Biostatistician YCAS geliang.gan@yale.edu

Contents Ø Why did I develop the macro Ø How did I develop the macro Ø What can the macro do for you Ø Limitation of the macro

Developing Table 1 macro Ø 2 macros were developed, one for categorical variable summary and one for continuous variable summary Ø 2 macros were combined into 1 overall table 1 macro % Table1Macro (dsn = car1, calist = type origin gender carmake, coplist = msrp invoice, cononplist = mpg_city mpg_highway, group = drivetrain);

Sample Result Table1 - Variable Summary Drivetrain Front Rear N = 336 P Value (N = 226) (N = 110) MSRP Mean (SD) 24782.56 (10697.04) 46093.63 (27306.18) <0.001 *** Invoice Mean (SD) 22834.07 (9718.06) 42101.36 (24925.46) <0.001 *** MPG (City) Median (Range) 21.0 (12.0 – 60.0) 18.0 (12.0 – 29.0) <0.001 *** MPG (Highway) Median (Range) 29.0 (18.0 – 66.0) 25.0 (18.0 – 36.0) <0.001 *** Type Hybrid 003 (01.33%) 000 (00.00%) <0.001 *** SUV 022 (09.73%) 000 (00.00%) Sedan 179 (79.20%) 055 (50.00%) Sports 008 (03.54%) 036 (32.73%) Truck 000 (00.00%) 012 (10.91%) Wagon 014 (06.19%) 007 (06.36%) Origin Asia 099 (43.81%) 025 (22.73%) <0.001 *** Europe 037 (16.37%) 050 (45.45%) USA 090 (39.82%) 035 (31.82%) gender 0.023 * * 0 098 (49.25%) 069 (62.73%) 1 101 (50.75%) 041 (37.27%) carmake 0 034 (15.04%) 023 (20.91%) 0.17 1 096 (42.48%) 051 (46.36%) 2 096 (42.48%) 036 (32.73%) Note: * P value less than 0.05 ** P value less than 0.01 *** P value less than 0.001

Macro parameters dsn = calist = coplist = cononplist = outputorder = continuousorder = 1 outdsn = eratio = 0.2 percenttype = copmain = mean table1summary column copsupplement = cononpmain = cononpsupplement caexcludelist = copexcludelist = std median = range cononpexcludelist = caincludemissing = copincludemissing = cononpincludemissi missingtop = yes no no ng = no notest = no cadec = 2 copdec = 2 cononpdec = 1 overall = no paddingchar = 1 reportinsas = no variableshading = labelvariable = no suppresswarning = yes yes creatertf = yes compacttable = yes tabletitle = savefilename = orderbyformat = variable summary pvaluetop = no orientation = missingpercent = missinglabel = group = yes Missing showgroup = yes Nalabel = NA

Foolproof features • dsn must be provided and data isn’t empty • Variable name in any list must be in the dsn • Name of statistic for continuous variable is the list supported by proc means • Stats are only calculated once for duplicate variable names in the list • If group variable are all missing, data will be summarized without stratification • Parameter eratio ranges between 0 and 1 • Variable in continuous variable list must be numeric

Example1: overall = yes % Table1Macro (dsn = car1, calist = type origin gender carmake, coplist = msrp invoice, cononplist = mpg_city mpg_highway, group = drivetrain, Overall = yes );

Example1 Result Table1 - Variable Summary Drivetrain Front Rear Total P Value (N = 226) (N = 110) (N = 336) MSRP Mean (SD) 24782.56 (10697.04) 46093.63 (27306.18) 31759.40 (20488.22) <0.001 *** Invoice Mean (SD) 22834.07 (9718.06) 42101.36 (24925.46) 29141.82 (18643.15) <0.001 *** MPG (City) Median (Range) 21.0 (12.0 – 60.0) 18.0 (12.0 – 29.0) 20.0 (12.0 – 60.0) <0.001 *** MPG (Highway) Median (Range) 29.0 (18.0 – 66.0) 25.0 (18.0 – 36.0) 28.0 (18.0 – 66.0) <0.001 *** Type <0.001 *** Hybrid 003 (01.33%) 000 (00.00%) 003 (00.89%) SUV 022 (09.73%) 000 (00.00%) 022 (06.55%) Sedan 179 (79.20%) 055 (50.00%) 234 (69.64%) Sports 008 (03.54%) 036 (32.73%) 044 (13.10%) Truck 000 (00.00%) 012 (10.91%) 012 (03.57%) Wagon 014 (06.19%) 007 (06.36%) 021 (06.25%) Origin Asia 099 (43.81%) 025 (22.73%) 124 (36.90%) <0.001 *** Europe 037 (16.37%) 050 (45.45%) 087 (25.89%) USA 090 (39.82%) 035 (31.82%) 125 (37.20%) gender 0 098 (49.25%) 069 (62.73%) 167 (54.05%) 0.023 * * 1 101 (50.75%) 041 (37.27%) 142 (45.95%) carmake 0 034 (15.04%) 023 (20.91%) 057 (16.96%) 0.17 1 096 (42.48%) 051 (46.36%) 147 (43.75%) 2 096 (42.48%) 036 (32.73%) 132 (39.29%) Note: * P value less than 0.05 ** P value less than 0.01 *** P value less than 0.001

Example2: show missing information % Table1Macro (dsn = car1, calist = type origin gender carmake, coplist = msrp invoice, cononplist = mpg_city mpg_highway, group = drivetrain, Overall = yes, caincludemissing = yes, copincludemissing = yes, cononpincludemissing = yes );

Example2 Result Table1 - Variable Summary Drivetrain Front Rear Total P Value (N = 226) (N = 110) (N = 336) MSRP N (N Missing) 226 (0) 110 (0) 336 (0) Mean (SD) 24782.56 (10697.04) 46093.63 (27306.18) 31759.40 (20488.22) <0.001 *** Invoice N (N Missing) 226 (0) 110 (0) 336 (0) Mean (SD) 22834.07 (9718.06) 42101.36 (24925.46) 29141.82 (18643.15) <0.001 *** MPG (City) N (N Missing) 213 (13) 100 (10) 313 (23) Median (Range) 21.0 (12.0 – 60.0) 18.0 (12.0 – 29.0) 20.0 (12.0 – 60.0) <0.001 *** MPG (Highway) N (N Missing) 221 (5) 107 (3) 328 (8) Median (Range) 29.0 (18.0 – 66.0) 25.0 (18.0 – 36.0) 28.0 (18.0 – 66.0) <0.001 *** Type Missing 000 (00.00%) 000 (00.00%) 000 (00.00%) <0.001 *** Hybrid 003 (01.33%) 000 (00.00%) 003 (00.89%) SUV 022 (09.73%) 000 (00.00%) 022 (06.55%) Sedan 179 (79.20%) 055 (50.00%) 234 (69.64%) Sports 008 (03.54%) 036 (32.73%) 044 (13.10%) Truck 000 (00.00%) 012 (10.91%) 012 (03.57%) Wagon 014 (06.19%) 007 (06.36%) 021 (06.25%) Origin Missing 000 (00.00%) 000 (00.00%) 000 (00.00%) <0.001 *** Asia 099 (43.81%) 025 (22.73%) 124 (36.90%) Europe 037 (16.37%) 050 (45.45%) 087 (25.89%) USA 090 (39.82%) 035 (31.82%) 125 (37.20%) gender Missing 027 (11.95%) 000 (00.00%) 027 (08.04%) 0.023 * * 0 098 (43.36%) 069 (62.73%) 167 (49.70%) 1 101 (44.69%) 041 (37.27%) 142 (42.26%) carmake Missing 000 (00.00%) 000 (00.00%) 000 (00.00%) 0.17

Example3: exclude missing row in percentage calculation % Table1Macro (dsn = car1, calist = type origin gender carmake, coplist = msrp invoice, cononplist = mpg_city mpg_highway, group = drivetrain, Overall = yes, caincludemissing = yes, copincludemissing = yes, cononpincludemissing = yes, missingpercent = no );

Example3 Result Table1 - Variable Summary Drivetrain Front Rear Total P Value (N = 226) (N = 110) (N = 336) MSRP N (N Missing) 226 (0) 110 (0) 336 (0) <0.001 *** Mean (SD) 24782.56 (10697.04) 46093.63 (27306.18) 31759.40 (20488.22) Invoice N (N Missing) 226 (0) 110 (0) 336 (0) Mean (SD) 22834.07 (9718.06) 42101.36 (24925.46) 29141.82 (18643.15) <0.001 *** MPG (City) N (N Missing) 213 (13) 100 (10) 313 (23) Median (Range) 21.0 (12.0 – 60.0) 18.0 (12.0 – 29.0) 20.0 (12.0 – 60.0) <0.001 *** MPG (Highway) N (N Missing) 221 (5) 107 (3) 328 (8) Median (Range) 29.0 (18.0 – 66.0) 25.0 (18.0 – 36.0) 28.0 (18.0 – 66.0) <0.001 *** Type <0.001 *** Missing 000 (00.00%) 000 (00.00%) 000 (00.00%) Hybrid 003 (01.33%) 000 (00.00%) 003 (00.89%) SUV 022 (09.73%) 000 (00.00%) 022 (06.55%) Sedan 179 (79.20%) 055 (50.00%) 234 (69.64%) Sports 008 (03.54%) 036 (32.73%) 044 (13.10%) Truck 000 (00.00%) 012 (10.91%) 012 (03.57%) Wagon 014 (06.19%) 007 (06.36%) 021 (06.25%) Origin Missing 000 (00.00%) 000 (00.00%) 000 (00.00%) <0.001 *** Asia 099 (43.81%) 025 (22.73%) 124 (36.90%) Europe 037 (16.37%) 050 (45.45%) 087 (25.89%) USA 090 (39.82%) 035 (31.82%) 125 (37.20%) gender Missing 027 (11.95%) 000 (00.00%) 027 (08.04%) 0.023 * * 0 098 (49.25%) 069 (62.73%) 167 (54.05%) 1 101 (50.75%) 041 (37.27%) 142 (45.95%) carmake Missing 000 (00.00%) 000 (00.00%) 000 (00.00%) 0.17

Limitations • Program is developed using SAS 9.4.3 and only fully tested in Windows 7 and Word 2013 • Don’t change the background color of your document • If missing data are labeled with special values, such as “999”, and there are true missing, the report may have unexpected errors • Program may halt in calculating some p value for Fisher’s Exact test

A million thanks to my colleague at YCAS for their support on developing this macro geliang.gan@yale.edu

A Macro Automates your Table 1 Geliang Gan Biostatistician YCAS - PowerPoint PPT Presentation

Yale Center for Analytical Sciences (YCAS) A Macro Automates your Table 1 Geliang Gan Biostatistician YCAS geliang.gan@yale.edu Contents Why did I develop the macro How did I develop the macro What can the macro do for you

Q1,2020 FIRST INVESTMENT BANK CONFIDENTIAL Table of contents Section 1 Macro environment 3

Q4 2017/ FY2017 Earnings Call February 2018 1 Table of Content Indonesia Macro Overview

Automates Cellulaires P . Guillon et G. Theyssier quipe GDAC, I2M nov. 2017 Plan de

ANSIBLE BEST PRACTICES: THE ESSENTIALS Ansible Automates: DC Jamie Duncan @jamieeduncan

When Inequality Matters for Macro and Macro Matters for Inequality Discussion by Christopher

August 7 2019 www.firstbanknigeria.com Outline Macro, T rending News..Global Macro, T

Art Center Camera Club February Workshop Macro Photography MACRO PHOTOGRAPHY by......Pete

Macro and Close up Photography What does Macro mean? Extreme close-up photography, usually of

Nautilus Automates, You Analyze Reproducible Analysis of Sensor Data in R Dr. Benjamin Ashwell

MACRO GROUP | PRESENTATION MACRO GROUP | SHAREHOLDERS Ctia Rebelo, Financial, HR & Quality

PCORI Online Training for Pre-Award Management System April 2017 Why Change? Automates and

Shocks and Propagation in Traditional and Modern Macro Robert J. Gordon Northwestern University,

Macro Photography Macro Photography Photography of small items larger than life size. Typical

Macro Photography Macro Photography Photography of small items larger than life size. Typical

Scripting using ImageJ Macro - a quick 1 hour tutorial - We use Fiji. http://fiji.sc please

A macro is user-defined syntactic sugar. A macro definition

Electric Vehicle Charger Placement Project NAME M I C H I G A N S T A T E U N I

Macro-Industrial Working Group Meeting 2: Industrial updates and Preliminary results Macro

2020 Forward-Looking Statements This presentation includes forward-looking statements, as

Group Brazil Back to the game? Brazil Back to the game? MACRO FUNDAMENTALS MONETARY POLICY

Recent macroeconomic trends in emerging economies and implications for development: A Case of India

Using Comparative Inventory to Bet Against the Oil Market Art Berman Labyrinth Consulting

for IEEE standards drafts Lisa Perry IEEE-SA Content Production and Management Topics XML

Person-Centred Practice Professor Tanya McCance Director of Institute of Nursing and Health

A Macro Automates your Table 1 Geliang Gan Biostatistician YCAS - PowerPoint PPT Presentation

Yale Center for Analytical Sciences (YCAS) A Macro Automates your Table 1 Geliang Gan Biostatistician YCAS geliang.gan@yale.edu Contents Why did I develop the macro How did I develop the macro What can the macro do for you

Q1,2020 FIRST INVESTMENT BANK CONFIDENTIAL Table of contents Section 1 Macro environment 3

Q4 2017/ FY2017 Earnings Call February 2018 1 Table of Content Indonesia Macro Overview

Automates Cellulaires P . Guillon et G. Theyssier quipe GDAC, I2M nov. 2017 Plan de

ANSIBLE BEST PRACTICES: THE ESSENTIALS Ansible Automates: DC Jamie Duncan @jamieeduncan

When Inequality Matters for Macro and Macro Matters for Inequality Discussion by Christopher

August 7 2019 www.firstbanknigeria.com Outline Macro, T rending News..Global Macro, T

Art Center Camera Club February Workshop Macro Photography MACRO PHOTOGRAPHY by......Pete

Macro and Close up Photography What does Macro mean? Extreme close-up photography, usually of

Nautilus Automates, You Analyze Reproducible Analysis of Sensor Data in R Dr. Benjamin Ashwell

MACRO GROUP | PRESENTATION MACRO GROUP | SHAREHOLDERS Ctia Rebelo, Financial, HR &amp; Quality

PCORI Online Training for Pre-Award Management System April 2017 Why Change? Automates and

Shocks and Propagation in Traditional and Modern Macro Robert J. Gordon Northwestern University,

Macro Photography Macro Photography Photography of small items larger than life size. Typical

Macro Photography Macro Photography Photography of small items larger than life size. Typical

Scripting using ImageJ Macro - a quick 1 hour tutorial - We use Fiji. http://fiji.sc please

A macro is user-defined syntactic sugar. A macro definition

Electric Vehicle Charger Placement Project NAME M I C H I G A N S T A T E U N I

Macro-Industrial Working Group Meeting 2: Industrial updates and Preliminary results Macro

2020 Forward-Looking Statements This presentation includes forward-looking statements, as

Group Brazil Back to the game? Brazil Back to the game? MACRO FUNDAMENTALS MONETARY POLICY

Recent macroeconomic trends in emerging economies and implications for development: A Case of India

Using Comparative Inventory to Bet Against the Oil Market Art Berman Labyrinth Consulting

for IEEE standards drafts Lisa Perry IEEE-SA Content Production and Management Topics XML

Person-Centred Practice Professor Tanya McCance Director of Institute of Nursing and Health

MACRO GROUP | PRESENTATION MACRO GROUP | SHAREHOLDERS Ctia Rebelo, Financial, HR & Quality