I have a SAS file which is opened under SERVER=SHARE, because it can be used by multiple users.  However, for the majority of users, we want to disable the DELETE command, so they can view / Add / Dup / MOD, but they will NOT be able to delete ay record.    And for some power users, we want to keep the DELETE command on and operational.   Can this be done using SCL code  ?   I have tried to use a special "PROC FSEDIT DATA=dataset_name NOBORDER NODELETE;" for them,  and a standard PROC FSEDIT DATA=dataset_name NOBORDER; for the other users, but as soon as the first one (with teh NODELETE) is used once, the whole dataset becomes in a "NODELETE" for everyone 😞     Thanks for giving me any way to have SAS makes my expectations come true.   Sylvain   ... View more

Hi, We are looking to improve data inserts in a SQL Server database (ODBC). In the SAS server connection, we have the advanced option to set 'Bulk Load' to Yes:     but we can also set this advanced option in the ODBC setup.   Do we have to set it in both?   Thx. ... View more

Dear SAS Community,   I get this 'ERROR: All variables in array list must be the same type' when using this code after transposing data from wide to long format. All this variables (FruitWt, Color, Rind, Button, Wtloss) are numeric variables, but since my data set starts with missing values (.) maybe that is why? I would greatly appreciate your help!   data long;     set new;     array _FruitWt (*) Day0_FruitWt Day3_FruitWt Day7_FruitWt ;     array _Color(*) Day0_Color Day3_Color Day7_Color ; array _Rind (*) Day0_Rind Day3_Rind Day7_Rind; array _Button (*) Day0_Button Day3_Button Day7_Button; array _Wtloss (*) Day0_Wtloss Day3_Wtloss Day7_Wtloss;     do i=1 to dim(_FruitWt);         label=scan(vname(_FruitWt(i)), 1, "_");         FruitWt=_FruitWt(i);         Color=_Color(i); Rind=_Rind(i); Button=_Button(i); Wtloss=_Wtloss(i);         output;     end;     drop Day0_: Day3_: Day7_: i; run;   Here is the data new:     Obs Test Grower Treatment FruitNr Day0_FruitWt Day0_Color Day0_Rind Day0_Button TRT_day3 Day3_FruitWt Day3_Color Day3_Rind Day3_Button Day7_FruitWt Day7_Color Day7_Rind Day7_Button Day3_Wtloss Day7_Wtloss 1 1 1 Initial 1 146.15 6 0 0 NA . . . . . . . . . . 2 1 1 Initial 2 218.42 5 0 3 NA . . . . . . . . . . 3 1 1 Initial 3 164.77 5 0 3 NA . . . . . . . . . . 4 1 1 Initial 4 174.28 6 0 3 NA . . . . . . . . . . 5 1 1 Initial 5 157.36 6 0 3 NA . . . . . . . . . . ... View more

I asked a similar question in a different thread for a single variable proc freq table here, but was curious if it could also be done with a multi-variable analysis.   Essentially, the title of the thread: is it possible to order a column in a multi-variable proc freq table by a separate variable not being analyzed? For example, take the following data:   County_Name Yes_No County_Code A Yes 1 A Yes 1 A Yes 1 A No 1 B No 2 B No 2 B No 2 C Yes 5 C Yes 5 C Yes 5 C Yes 5 C No 5 D No 4 D No 4 D Yes 4 E Yes 3 E Yes 3 E No 3 E No 3 F Yes 6 F Yes 6 F No 6 F No 6   Normally, if I made a proc freq table to count the frequency of yes/no's for each county, as so: proc freq data=data; tables County_Name*Yes_No; run;    I would get an output table similar to the following: County_Name Freq       Yes No Total A 3 1 4 B 0 3 3 C 4 1 5 D 1 2 3 E 2 2 4 F 2 2 4   However, is it possible to produce a similar table, but have the County_Name variable ordered according to its corresponding County_Code value? So, a table that would look like this instead: County_Name Freq       Yes No Total A 3 1 4 B 0 3 3 E 2 2 4 D 1 2 3 C 4 1 5 F 2 2 4 I tried to do something similar to the single-variable solution, but instead got a larger cross-tabulation that gave a separate row for each combination, instead of having the yes/no frequency in their own columns. Something like this:   proc freq data=data; tables County_Code*County_Name*Yes_No / list; run;   County_Code County_Name Yes_No Freq 1 A Yes 3 1 A No 1 2 B Yes 0 2 B No 3 3 C Yes 4 3 C No 1 4 D Yes 1 4 D No 2 5 E Yes 2 5 E No 2 6 F Yes 2 6 F No 2   ... View more