SAS PROGRAMMING CS 2331
Popular in Course
Popular in ComputerScienence
This 0 page Class Notes was uploaded by Ms. Bridie Kohler on Sunday November 1, 2015. The Class Notes belongs to CS 2331 at Oklahoma State University taught by Staff in Fall. Since its upload, it has received 24 views. For similar materials see /class/232839/cs-2331-oklahoma-state-university in ComputerScienence at Oklahoma State University.
Reviews for SAS PROGRAMMING
Report this Material
What is Karma?
Karma is the currency of StudySoup.
You can buy or earn more Karma at anytime and redeem it for class notes, study guides, flashcards, and more!
Date Created: 11/01/15
PROC GPLOTS You say plots earlier The plots have a few new features to examine These are the symbol interpol and line symbol is used to de ne points on the plot with a certain symbol These can vary interpol or I is used to join the symbols together There are a few options available These arejoin spline rl or clm interpolj0in means that the points are joined together interpolspline makes a smooth line through the points interpolrl is used for regression analysis This nds the best t for a line through the plot The 1 stands for regression line interpolclm is used to de ne con dence limits for a mean predicted value These are usually set to a 95 percent limit line or L allows you to specify what type of line you want to use There are 46 different line styles line1 would give you a solid line You can add width to a line to make it thicker Height can also be used in a SYMBOL statement You can emphasize the symbol if you want This rst plot is a line chart It displays the same information as the hbar chart of Comparison of Undergraduate Enrollments There are three lines with a symbol that represents the total number of students for each year One of the reasons that this particular plot is shown here is that there is more than one way to display the same information It would depend on your audience or what you think gets your point across the best You would have to decide on which one you like the best data maj or3 set major symboll vstar h2 ispline l1 cred symbolZ Vcircle h2 ispline l10 cgreen symbol3 vdiam0nd h2 ispline l15 cblue proc gplot where college 39AS or college 39BU or college EN39 plot totalquot yearcollege format college univ title 39Plot of College Enrollments by Year at OSU39 237 title2 39Comparison of Undergraduate Enrollments for 1997to 2002 footnotejl Source OSU Student Pro le Fall 2002 footnote2 fswissb Output 3139 run These next plots give examples of how you can plot out data but the picture does not help you see much of a picture or trend But by using some other features you can make the plot much more readable and understandable This particular graph contains blood pressure readings for one month It contains the date which contains the month day and year which is entered in a four year digit It contains a reading for morning and one for the evening for each day of the month It has the systolic and diastolic pressure readings The data looks similar to this 12092000 AM 152 68 12092000 PM 121 66 12102000 AM 15167 12102000 PM 120 64 The input statement for this data could be input 1 date mmddyy8 time 35 10 11 systolic 13 15 diastolic 16 17 When the data is read in you can use one of the date informat statements The date could be read in with an mmddyy8 informat When the plot is developed the Systolic pressure will be on the Vertical Axis and the Day of the Month will be on the horizontal axis One date format that could be used to keep the dates short would be day2 This is not a date format that you used before but it will list out the day ofthe month 1 2 3 25 26 etc However best laid plans do not always work out the way we want The dates would not list out the way I planned on the horizontal axis So if at first you don t succeed try some other approach This is the input statement that I used to make these plots input month 1 2 day 3 4 year 5 8 time 10 11 systolic 13 15 diastolic 17 18 This particular input statement left more exibility for the plots goptions border goptions ftitle swiss ftextzapf 238 OUTPUT 3 1 data bloodp in le csas le2bloodp2000dat input month 12 day 34 year 58 time 1011 systolic 1315 diastolic 1718 239 label month Month day 39Day of Month year 39Year39 time 39Time of Day systolic 39Systolic39 diastolic 39Diastolic39 data decemberall set bloodp if month 12 symbol Vstar h2 c blue proc gplot plot systolicday title cgreen 39December Blood Pressure Readings footnote fswissb Output 32 run This particular plot uses the default aXis measurements The days of the month are set up by SAS This plot shows all of the blood pressure readings from both morning and evening in the month of December The graph is difficult to look at and get any pattern that would be helpful Before looking at some other ways to make the graph easier to understand look at an example with a Regression line through the plot Regression deals with finding the equation of a straight line that best fits the points on a plot data decemberall set bloodp if month 12 symbol Vstar h2 c blue ir proc gplot plot systolicday title cgreen 39December Blood Pressure Readings title2 Use ofa Regression Line footnote fswissb Output 33 run The I is for Interpol joining points together In this case the r is used for regression You will notice a line through the graph 240 OUTPUT 3 2 241 OUTPUT 3 3 These are some features that can be added to make the plot more understandable First of all one item is to put some vertical reference lines in the graph to show where high blood pressure is and then where the borderline pressure is and where is normal pressure These are looking at systolic pressure Some medical standards de ne blood pressure as high if it is 140 or over Borderline blood pressure is considered 130 to 139 You will see a cvref added also It 242 allows you to make the vertical reference lines a different color We can specify what the vertical axix and horizontal axis contain The vaxix 90 to 170 by 10 for the vertical axis will be used For the horizontal axis haxis l to 31 by 2 will by used The next two graphs will also split up the number of symbols by only using the AM data on one graph and the PM data on a separate graph Lines will be connected to the symbols also to help us understand the graphs This is the program data december set blood if month 12 if time AM symbol fmarker vU h1 a blue ispline l2 w2 proc gplot plot systolicdayvref 130 140 cvrefgreen vaxis 90 to 170 by 10 haxis 1 to 31 by 2 title cgreen 39Morning December Blood Pressure39 footnote fswissb 39Output 3439 run Notice that there are some new types of symbols These are all filled in with a solid color They are a different type of font Some of them will be listed at the end of this explanation Next here is a graph for the PM blood pressure readings data december2 if time PM symbol fmarker vC cgreen ispline l6 w4 proc gplot plot systolicdayvref 130 140 cvrefblue vaxis 90 to 170 by 10 haxis 1 to 31 by 2 title cblue 39Evening December Blood Pressure39 footnote fswissb 39Output 3539 run These two graphs make it easier to tell when the blood pressure readings were too high or when they were in the normal range The days of the month across the horizontal axis made it easier to see the dates The next graph will show both the morning and evening readings on the same graph 243 OUTPUT 34 244 OUTPUT 35 data december3 if month 12 symboll fontmarker valueU cb1ue ispline 16 w4 symbolz font marker vC cgreen ispline 19 w4 proc gplot plot systolicdaytimevref 130 140 vaxis 90 to 13970 by 10 245 haxis 1 to 31 by 2 title 39Morning amp Evening December Blood Pressure39 footnote fswiss 39Output 3639 run Although all of the data listed on this graph is the same as in the first graph on the blood pressure example it is easier to see and you can tell which ones are too high and which ones are more in the normal range Some different symbols were used in these graphs These are from a font series called marker The symbols that are shown in this font are colored in and are larger than the regular symbols These are F marker V U Square filled in F marker V P Diamond filled in F marker V C Triangle filled in F marker V V Star filled in F marker V Q Figure ofa man F marker V R Figure ofa woman If you want to use some of these it is better to use either all of them where you just the regular symbols or use the marker symbols They don t always mix well when you try to mix them together The next plot displays the total number of students by year A different symbol is used and the interpol is join in this case Join connects the points together A line selection is used also data project infile aproj ectdat input year 14 total 610 label total Number of Graduates symboll Vd0t cred ij0in l2 proc gplot plot t0talyear frame title High School Graduates in Public Schools in Oklahoma title2 Based on Current Public School Enrollment footnote j 1 Source OSU Student Pro le Fall 2002 footnote2 fswissb Output 37 run 246 OUTPUT 36 OUTPUT 3 7 247 This next plot will use some other features The horizontal and vertical axes have been changed and there is also an hminor and a vminor used These allow you to cut down the number of minor tick marks on the axis that are between the major numbers The autovref will add grid lines to the plot at every major tick mark on the vertical axis These make it easy to see where the numbers cross The color grayaa is used to make a gray line that is lighter to make the grid contrast more If you wanted the reference lines at certain horizontal lines you could specify these years An example would be hrefl994 1996 1998 Lines would be drawn at these three places 248 data proj2 set project symboll value diamond h2 crred ispline 12 plot t0talyear frame haxis 1993 to 2007 by 2 hmin0r1 vaxis 30000 to 40000 by 1000 vminor 1 autovref cvrefgrayaa title High School Graduates in Public Schools in Oklahoma title2 Based on Current Public School Enrollment footnote j 1 Source OSU Student Pro le Fall 2002 footnote2 fswissb Output 38 run This last plot used an area statement The area will ll in part of the chart under the line It uses an AREA statement A symbol for a line is used here and then a pattern statement is used to de ne how the area will be lled in under the line See the following program data proj3 set project symboll value none ij0in pattern vs cred plot totalyear haXis 1993 to 2007 by 2 vaXis 30000 to 40000 by 1000 areas 1 title High School Graduates in Public Schools in Oklahoma title2 Based on Current Public School Enrollment footnote j 1 Source OSU Student Pro le Fall 2002 footnote2 fswissb Output 39 run 249 OUTPUT 38 250 OUTPUT 39 SASGRAPH also has a 3D feature that allows you to make graphs on an X and y and z axis These can be used for mathematical formulas You can change the angle and rotation of the design SASGRAPH has other features such as the GMAP feature which allows you to make maps of the entire United States and also of each state in the US The states are broken down into counties There are also maps of the world and several countries There are many features to SASGRAPH that are not shown here You have just scratched the surface in this class If you are interested there are many books that you can 251 obtain to learn more about SAS GRAPH and SAS This is just any extra note In order to get the graphs to print out for your notes I had to use some features in the PRINT menu I had to use PRINT PAGE SETUP I changed the orientation from Portrait to Landscape I also changed the margins to 1 top bottom left and right 252 PROC GCHARTS Bar Charts Bar charts are used frequently You have been introduced to bar charts already There are some special features that can be used with PROC GCHARTS The main features are PATTERN and VALUE statements You will also see many other features that you can use When you make a graph SAS will ll in the bars with default designsusually cross hatching SAS may also choose colors if none are listed If you want some control over the patterns and colors in the bars you can use the PATTERN statement It consists of the PATTERN statement and the VALUE statement Each of these graphs will use the same font for titles and for all other text on the charts It is important to have good readable text that is large and clear and uncluttered You can define them at the beginning of your program by using a GOPTIONS statement The GOPTIONS is a global statement and will be used for all graphs unless you change it in the program All of the programs used in these examples use Goptions ftitle swiss ftext zapf Goptions border Goptions rotate landscape The ftitle specifies what the font will be for the title in each graph The ftext specifies what all other text will be in the graph This means the axis text labels and legends will use this second font The font swiss is a clear font and is dark and will stand out in the titles The font zapf is a darker font and shows up better on graphs This is helpful when you are making overheads as the print stands out more The border option is used to put a border around the entire graph to make it look better for the graphs in this book The graphs will also be printed out in a landscape form All data used for these graphs will be at the end of the unit on SASGRAPHS Examples will start with a simple graph More features will be added as you go along Here is the rst example data all infile ac0112002dat39 input year 14 college 67 total 912 label total 39Total Students39 college College if year 2002 pattern valuesolid cred proc gchart 210 Vbar college sumvar total noframe title cred 39OSU College Enrollments title2 cblack 39Fall 200239 footnote cblackjleft 39Source OSU Student Pro le Fall 200239 footnote2 fswissb Output 2139 run Notice that after the 39vbar college39 the 39sumvar otal was used In this particular case only the summary data was availablenot an entire data set The sumvar statement had to be used here All of these graphs used summary data They would all contain the sumvar otal The sumvar otal is only used in this first chart so you can see how to work with summary data If you had a regular large data set you would not use the sumvar statement The line would read quotvbar collegequot All other examples will assume that you have an entire data set The noframe was used in this example because SAS puts a default frame around the graph itself where the bars are shown if you don t specify the noframe You will see the frame in another example Pattern statements should always go before the proc gchart statement Since there is only one Pattern statement here all of the bars will have the same pattern The quotvaluesolidquot means that the bar will be completely filled in You can use V for short The values that are acceptable are 2 L1 L5 3 R1 R5 9399 mm The X stands for cross hatching The 1 to 5 explain how thick you want the fill The quot1quot is the thinnest and the quot5quot is the thickest The L stands for parallel lines that go to the LEFT The R stands for parallel lines that go to the RIGHT The S means that the bar will be filled solid The E means that the bar will remain empty You can use several pattern statements and also add a color statement This next example uses different patterns and also subgroups the bars by class Some other features in this program are the use of the comma in the total number of students The proc format statement is used to change the class variable from numbers to word descriptions A box is also used around the legend the definition of class in this case data two infile aclass2002dat39 211 input college 12 class 5 total 710 OUTPUT 21 SIMPLE VBAR CHART 212 label total Total Students class Class college College proc format value numbers 1 39Freshmen39 2 39Sophomores39 3 39Juniors39 4 39Seniors39 patternl wx3 cred pattern2 wr4 cblue pattern3 VX5 cgreen pattern4 W15 cmagenta legend frame label none proc gchart Vbar collegesubgroupclass noframe legend legend format class numbers total comma7 title crred 39OSU Enrollments by Class for Fall 200239 footnote jl 39Source OSU Student Pro le Fall 200239 footnote2 fswissb Output 2239 run There is also a 3D feature that will allow you to make a 3d chart This next example uses the same information as in the example above except for the vbar3d statement Here is the program patternl VX3 cred pattern2 Vr3 cblue pattern3 VX5 cgreen pattern4 V15 cmagenta legend frame label none proc gchart vbar3d collegesubgroupclass noframe legend legend format class numbers total comma7 title crred 39OSU Enrollments by Class for Fall 200239 footnote jl 39Source OSU Student Profile Fall 200239 footnote2 fswissb Output 23 run Note that the 3D chart did not use the patterns that were provided It did use the colors that were listed Apparently the 3D chart uses the solid pattern on charts This 3D feature is fairly new and I have not seen any information on it in any of the SAS GRAPH books I have 213 OUTPUT 22 SUBGROUPED CHART 214 OUTPUT 23 3D CHART 215 Part of learning to make graphs is to also learn about extra features that can be added to graphs to incorporate extra information This next example adds a few new features to the graph that was developed above These are the use of a frame around the graph itself The other is the use of the sum statement where the numbers appear above each of the bars The bars are also in descending order The comma feature is also being used If you had many bars you could not use the comma feature as the graph would be too crowded Look at the differences in these graphs patteml vx3 cred pattem2 vr3 cblue pattem3 vx5 cgreen pattem4 v15 cmagenta proc gchart vbar collegesubgroupclass sum descending format class numbers total comma6 title fswiss cred 39OSU Enrollments by Class for Fall 200239 footnote jl fswiss 39Source OSU Student Pro le Fall 2002 footnote2 fswissb Output 2439 run The noframe option was deleted from this graph If you do not put it in the program it will automatically put a frame around the bars Notice that there is a frame around this graph NOTE If you try out the graph above and the numbers do not appear above the bars it is because the bars are not wide enough You can fix this by adding the WIDTH feature This allows you to set the width of the bars You can change it to different widths You may need to experiment with different widths to get the results you want Here is an example of the vbar statement with the WIDTH feature vbar quot 39 Wp video sum J 139 width 7 This next bar chart uses the group feature The proc format is used here to change the short names of the colleges to more meaningful names NOTE When you have a proc format statement that is character in this caseuniv you must use the sign when you are describing it in the format statement This particular program also uses the subgroup statement to change the colors of each year The original data set also had information on other colleges besides the three listed here A where statement is used to choose only those colleges that were wanted for the chart 216 OUTPUT 24 217 data major infile amaj ors2002dat input year l4 college 67 total 9 12 proc format value 35qu 39AS39 39Arts amp Sciences39 39BU39 39Business39 39EN39 39Engineering39 label total Total39 patternl Vr3 cred pattern2 VX2 cblue pattern3 VS cgreen pattern4 Vl4 ccyan pattern5 VX4 cblack pattern6 VS cmagenta proc gchart where college 39AS39 or college 39BU39 or college 39EN39 vbar year discrete groupcollege subgroupyear format college univ title Oklahoma State University39 title2 39Comparison of Undergraduate Enrollments for 1997 to 200239 footnote Source OSU Student Pro le Fall 2002 footnote2 fswissb Output 2539 run Next is a horizontal bar chart It has the same information as the one above only the bars are now horizontal Notice that the totals are out to the side of each bar The use of legend in a box is also used here One problem that you will encounter with several bars of different designs is that you can create an optical illusion that the bars are crooked This is particularly a problem with a small graph If this was in color you could make more of the bars a solid color and avoid this problem The hbar follows data maj or2 set major proc format value univ 39AS 39Arts amp Sciences 39BU Business EN Engineering patternl Vr3 cred pattern2 VX2 cblue pattern3 VS cgreen pattern4 VX4 ccyan 218 OUTPUT 25 219 OUTPUT 26 patternS Vr4 cblack pattern6 VS cmagenta legend frame label n0ne 220 proc gchart where college 39AS or college 39BU or college EN39 hbar year discrete groupcollege subgroupyear legendlegend format total c0mma7 college univ title Oklahoma State University39 title2 39Comparison of Undergraduate Enrollments for 1997 to 200239 footnote2 fswissb Output 2639 run NOTE Occasionally when you make a horizontal bar chart there does not seem to be much space between the groups of bars and it is difficult to see the breaks in the groups The graph above was grouped by college You can add more space between the groups by using the GSPACE feature This may happen more with the horizontal bar charts An example of its use follows hbar year discrete groupcollege subgroupyear gspace 2 legendlegend There are a few other features that you may want to try When we first discussed charts you had an example of data that had a column called Size This was character data Three different bars were described Short Average and Tall When SAS displays them it puts them in alphabetical orderiAverage Short and Tall If you want it in another order you can use the MIDPOINTS option You can also change the colors on the graph using a PATTERNID option This can be used with midpoints and also group Each group would be a different color Here is an example of a program that will use both of these features Data wtloss Set healthy Length size 7 If height lt 64 then size SHORT Ifheight gt 65 and height lt 70 then size AVERAGE Ifheight gt 70 then size TALL Label size SIZE Pattern VS crred Pattern2 VS cgreen Pattern3 VS cblue Proc gchart Vbar sizemidpoints SHORT AVERAGE TALL pattemidmidp0int noframe width10 Title USE OF MIDPOINTS TO CONTROL ORDER OF CHARACTER DATA Title2 USE OF PATTERNID TO CHANGE COLORS OF EACH BAR Footnote fswissb Output 27 The next one uses two colors for a group See the following program 221 Proc gchart Vbar sizemidpoints SHORT AVERAGE TALL group sex pattemidgroup noframe width8 Title GRAPH OF SIZE GROUPED BY SEX Title2 USE OF PATTERNID WITH GROUP Footnote fswissb Output 28 Run 222 OUTPUT 27 223 OUTPUT 28 Pie Charts Pie charts only use one column They emphasize the whole of that particular column Pie charts have some more options They also use various patterns These take into consideration different angles based on 360 degrees The VALUE pattern can use the following 224 1 PEMPTY or PE 2 PSOLID 0r PS 3 P densityltstyleltanglegtgt where density can be 1 5 The thickness of the lines style can be X or N X is cross hatching N is parallel lines angle can be 0 360 the angles can be in degrees Some other features of the PIE are being able to change the location of the name of the variables and the amount of the section of the Pie type The default is to give freq statistics You can use typepercent to obtain the percentages Some other features are the slice and value fill explode and invisible slice can be used to place the slice labels inside the slices Use sliceinside If it is omitted the slice labels go outside value can be used with arrow This creates an arrow to connect the amounts with their corresponding slices ll allows you to fill in the slices with Solid The fillsolid will fill in each slice with default colors from SAS You can also use llX for cross hatching explode can be used to offset one particular slice invisible allows you to leave one of the slices empty across allows you to put two or three pies across a page landscape format down allows you to put two or three pies down the page portrait format This first program uses different pattern statements and the valuearrow The names of the colleges are changed to spell out their longer names Unfortunately these truncate at 16 characters data degree infile 39adegree2002dat input college 12 degree 45 total 69 225 proc format value name AG Agriculture Sci AS Arts amp Sciences BU Business Admin ED Education EN Engineering HE Human Envir Sci patteml Wp4n90 cred pattem2 Wps ccyan pattern3 Wp3n90 cblack pattem4 Wps cmagenta patterns Wp4x0 cblue pattern6 Wps cyellow proc gchart where degree BA pie college sumvar t0tal value arrow noheading format college name title Degrees Granted at OSU for 20012002 title2 Undergraduate Students footnotejl Source OSU Student Pro le Fall 2002 footnote2 fswissb Output 29 run The pie chart also will allow you to make a 3D graph The same graph that was used above will be used for the 3D graph Notice that even though there were patterns used all of the pieces of the pie are lled in with solid colors One other feature is being used here It is MATCHCOLOR It will use the same color to print out the name and number of each piece of p1e patteml vp4n90 cred pattem2 VpS ccyan pattem3 vp3n90 cblack pattem4 VpS cmagenta 226 OUTPUT 29 pattern5 Vp4X0 cblue pattern6 VpS cgreen proc gchart where degree 39BA pie3D college sumvar otal matchcolor value arrow noheading format college name title Degrees Granted at OSU for 2001200239 title2 39Undergraduate Students footnotej 1 Source OSU Student Pro le Fall 2002 footnote2 fswissb Output 21039 227 run You can also display two or more pie charts on the same page You use the group and across2 command SAS also puts in its own heading on the graphs If you want that removed you can use noheading A program follows data degree2 set degree proc format value grad BA BACHELORS CO GRADUATE if degree 39MA39 or degree DO39 then delete patteml Vp4n90 cred pattem2 VpS ccyan pattem3 Vp3n90 cblack pattem4 VpS cmagenta pattem5 Vp4X0 cblue pattem6 VpS cgreen proc gchart pie college sumvart0tal explode 39AS39 slice outside group degree across 2 noheading format degree grad title cblue 39Degrees Granted at OSU for 20012002 title2 cgreen 39Undergraduate and Graduate footnote jl fswiss 39Source OSU Student Pro le Fall 200239 footnote2 fswissb Output 211 run OUTPUT 2 10 228 229 OUTPUT 21 1 Notice that there is a problem when the AS piece of pie is exploded The names of the pieces of the pie AS cut into the titles of the two pies Bachelors and Graduate One way to x this would be to delete the explode on the pie Another way is to change the angle of the pie By default the rst slice in a pie chart starts at the 3 o clock position Then the other pieces go counterclockwise You can change the starting position with the ANGLE option You can start the rst slice ofa pie at the top 12 o clock by using ANGLE90 The pie can use several angles 7 0 to 360 Pieces of the pie are always listed in either alphabetical order or numeric order counterclockwise Changing the angle of the pie in the last graph will allow you to t the slices on the graph without the name running into the label title See the following graph for an example 230 proc gchart pie college sumvar otal explode slice outside group degree VASV across 2 angle 45 noheading format degree grad title cblue 39Degrees Granted at OSU for 20012002 title2 cgreen 39Undergraduate and Graduate footnote jl fswiss 39Source OSU Student Pro le Fall 2002 footnote2 fswissb Output 212 run Instead of across you can use down The graphs will have two graphs one underneath the other one This works when you are using portrait to print out your graphs You can use both the across and the down on the same graph You would have four graphs on the same page Notice that when there are two graphs on a page you cannot use the label names for the columns If you tried to use the label names for the columns they would take up all the space and the pie itself would be very small Using gray shades for the graphs instead of color When you give a presentation to a group of people you will want to use graphs that are done in color However when you have them in a publication you cannot always use color since it would be too expensive to duplicate In your class notes for SAS I make the graphs in color But they are copied in black and white because it would be more expensive if I had the graphs duplicated in color for your notes If you have to publish something you need to use colors that will copy well One way to do this is to use gray colors SAS has several colors that can be used You may have to try out several colors to see what will work The pie chart is one 231 OUTPUT 2 12 problem if you use all solid colors in the pieces They may blend together when they are copied and you can t tell where one piece ends and another begins Using patterns between solid slices ofthe pie can help Some of the gray patterns are listed as Grayaa This is the darkest gray Graybb Graycc Graydd Grayee Grayff This is white 232 The colors in the middle get lighter as they go from aa to ff There is one new feature that has been added here It is the COUTLINE This will put a colored outline around each bar and it will be colored in black The reason for the use of the coutline is that the last color grayff will be white and you would not see any bar if you did not use the coutlineThe following is a bar chart that uses all ofthese pattern vs cgrayaa dark gray pattern2 vs cgraybb pattern3 vs cgraycc pattern4 vs cgraydd patterns vs cgrayee pattern6 v cgrayff white proc gchart Vbar collegesubgroupcollege sumvar total width6 noframe coutlineb1ack title cred 39OSU College Enrollments39 title2 cblack 39 Fall 2002 III footnote jl cblack 39 Source OSU Student Profile Fall 2002 footnote2 fswissb 39Output 21339 run The next one is a pie chart It will use gray colors that have numbers attached to them instead of alphabetic characters They range from 22 to 99 Try to put a dark one next to a light one so that there is contrast patternl vs cgray22 Darkest gray pattern2 vs cgray88 pattern3 vs cgray33 pattern4 vs cgray99 Lightest gray patterns vs cgray44 pattern6 vs cgray77 proc gchart where degree 39BA39 pie college sumvar total value arrow noheader format college name title 39Degrees Granted at OSU for 2001 2002 OUTPUT 2 13 233 title2 39Undergraduate Students39 footnote jl fswiss quotSource OSU Student Profile Fall 2002 footnote2 fswissb 39Output 21439 run39 Review of Features Used in Proc Gchart These are some features used in PROC GCHART Proc gchart Vbar column options also Vbar3d Hbar column options Pie column options also pie3d 234 Options for Vbars and Hbars Subgroup Stacked bars with different colors Group Side by side charts Discrete Used with numeric data to eliminate continuous numbers Pattern value Vary from LlL5 RlRS XlXS S E Parallel lines to Cross hatching to Solid to Empty Color or c color Sumvar Used for summary data when you don t have an entire data set Noframe Used when you don t want a frame around the graph 7 default is frame Sum Adds the sum above a bar in a bar chart Width Width makes bars wider Gspace Allows more space between groups in charts Descending Bar charts go in descending order iHighest to Lowest Ascending Bar charts go in ascending order 7 Lowest to Highest Legend frame Puts a frame around the legend in the graph Midpoints Used to change order of character data in a graph Pattemid Used to give different patterns to each bar Type Used for percent frequency mean OUTPUT 214 235 Options for Pie charts Pattern value Vary from PE Empty PS Solid Pd density ithickness of line Pds style can be XCrosshatching or NParallel lines Pdsa angle can be 0360 degrees example P5X45 slice Place slice labels inside pie or outside value arrow connects the amounts with the slices ll llSOlld lls in each pie slice with a SAS default color explode Offsets one ofthe slices ofpie Invisible Leaves one of the slices empty Across Allows you to put two or three pies across a page Down Allows you to put to or three pies down a page 236 Group Used to group pies across or down a page Noheading Leaves off the default heading on pie charts Matchcolor The color of the print matches the color of the slice of pie Angle Change the angle where the pie starts normally at 3 o clock position 237
Are you sure you want to buy this material for
You're already Subscribed!
Looks like you've already subscribed to StudySoup, you won't need to purchase another subscription to get this material. To access this material simply click 'View Full Document'