*Delete only containing a-zA-Z; if anyalpha(var)>0 then delete; *delete any non-numeric; if notdigit(var)>0 then delete;
Also there is a PRX solution, if you prefer that route, but the above are probably the easiest. There are a lot of ANY and NOT functions - take a look at the manual for more.
On Mon, Jan 26, 2009 at 1:27 PM, J M <jas...@ucla.edu> wrote: > How can delete observations containing characters (eg 92z89 or abcd)? > I just want to keep observations with numeric values. > Thanks.
One way to get this done is using PRXMATCH function. Data want; input value $; if prxmatch("/[a-zA-Z]/",value)=0; /* Searches for a pattern match and returns the position at which the pattern is found */ cards; 1023442 92z89 abcd 5231295 09CX42 98722 ;
On Mon, Jan 26, 2009 at 2:27 PM, J M <jas...@ucla.edu> wrote: > How can delete observations containing characters (eg 92z89 or abcd)? > I just want to keep observations with numeric values. > Thanks.
I do not weather you want to delete any Observation (Not a good Idea depending what the value is ) seems like you are working some data cleaning . In case you change and want to keep those observations too with numerical part here is way I think
Why don't you consider using New Compress function ( If you Using sas9 or above )
var = Compress ("Your variable", , "Kd") ;*Kd is keep digits I think *
Or if You are using SAS 8
Var = Compress ("Your Variable", "ABCDEFGHIJKLMNOPQRSTUVXYZ") * Hope I did not miss a letter*
I have to try Joe Matise method a try seems to easy off course there is PRX way but it is little bit confusing for beginners
On Mon, Jan 26, 2009 at 2:27 PM, J M <jas...@ucla.edu> wrote: > How can delete observations containing characters (eg 92z89 or abcd)? > I just want to keep observations with numeric values. > Thanks.
On Mon, 26 Jan 2009 11:27:55 -0800, J M <jas...@UCLA.EDU> wrote: >How can delete observations containing characters (eg 92z89 or abcd)? >I just want to keep observations with numeric values. >Thanks.
How about 12e3? It contains an alpha character but is nevertheless valid as the external representation of a numeric value?
Have to say, that's the first time I've come across ??. Is that the only place it works to suppress invalid messages, or does it work elsewhere? (For example, a proc transpose where you have columns overwriting other columns intentionally, etc.)?
-Joe
On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; <
Format Modifiers for Error Reporting ? suppresses printing the invalid data note when SAS encounters invalid data values.
See Also: How Invalid Data is Handled
?? suppresses printing the messages and the input lines when SAS encounters invalid data values. The automatic variable _ERROR_ is not set to 1 for the invalid observation.
They work on input statement and INPUT function but not inputN/C. As far as PROC TRANSPOSE goes perhaps LET will produce the effect you desire but I'm not sure exactly what you are referring to. Can you make example?
On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote:
> Have to say, that's the first time I've come across ??. Is that the only > place it works to suppress invalid messages, or does it work elsewhere? > (For example, a proc transpose where you have columns overwriting other > columns intentionally, etc.)?
> -Joe
> On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; < > iebup...@gmail.com> wrote:
> > It is a bit easier to quiet invalid data messages with ??
> > if missing(input(var,??best12.)) then delete;
> > On 1/26/09, Arthur Tabachneck <art...@netscape.net> wrote: > > > Jason,
> > > If Howard's concern is valid and you need to keep numbers stated in > > > scientific notation, then I'd suggest the following method:
> *Delete only containing a-zA-Z; > if anyalpha(var)>0 then delete; > *delete any non-numeric; > if notdigit(var)>0 then delete;
> Also there is a PRX solution, if you prefer that route, but the above are > probably the easiest. There are a lot of ANY and NOT functions - take a > look at the manual for more.
> -Joe
> On Mon, Jan 26, 2009 at 1:27 PM, J M <jas...@ucla.edu> wrote:
>> How can delete observations containing characters (eg 92z89 or abcd)? >> I just want to keep observations with numeric values. >> Thanks.
<iebup...@GMAIL.COM> wrote: >Format Modifiers for Error Reporting >? >suppresses printing the invalid data note when SAS encounters invalid >data values.
>See Also: How Invalid Data is Handled
>?? >suppresses printing the messages and the input lines when SAS >encounters invalid data values. The automatic variable _ERROR_ is not >set to 1 for the invalid observation.
>They work on input statement and INPUT function but not inputN/C. As >far as PROC TRANSPOSE goes perhaps LET will produce the effect you >desire but I'm not sure exactly what you are referring to. Can you >make example?
>On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: >> Have to say, that's the first time I've come across ??. Is that the only >> place it works to suppress invalid messages, or does it work elsewhere? >> (For example, a proc transpose where you have columns overwriting other >> columns intentionally, etc.)?
>> -Joe
>> On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; < >> iebup...@gmail.com> wrote:
>> > It is a bit easier to quiet invalid data messages with ??
>> > if missing(input(var,??best12.)) then delete;
>> > On 1/26/09, Arthur Tabachneck <art...@netscape.net> wrote: >> > > Jason,
>> > > If Howard's concern is valid and you need to keep numbers stated in >> > > scientific notation, then I'd suggest the following method:
HMMMM.... I think I need to write to Gmail team or request Howard which one would be easier ?? Hope Howard is reading this. I really miss his expert reply's may be all the Gmail users
thanks SL
On Mon, Jan 26, 2009 at 7:38 PM, ./ ADD NAME=Data _null_; <
iebup...@gmail.com> wrote: > No Howard's e-mail refuses to be accepted by gmail as anything but > spam. I can "STAR" them and that makes them easier to find in the > spam box.
> The only think I can think of is that JUNK in the e-mail address is > too much for G-Mail's spam filter to resist.
> I continue to check my spam box throughout the day and look forward to > seeing spam that has been starred.
> Thankfully e-mail from "Ajay who's it" does end up in the spam box as > directed.
> On 1/26/09, SAS_learner <procconte...@gmail.com> wrote: > > Hello _null_,
> > Did you figure out how to get Howard's email to Gmail Inbox. If so can > you > > let me know ??
> > Thanks > > SL
> > On Mon, Jan 26, 2009 at 5:48 PM, ./ ADD NAME=Data _null_; > > <iebup...@gmail.com> wrote: > > > Format Modifiers for Error Reporting > > > ? > > > suppresses printing the invalid data note when SAS encounters invalid > > > data values.
> > > See Also: How Invalid Data is Handled
> > > ?? > > > suppresses printing the messages and the input lines when SAS > > > encounters invalid data values. The automatic variable _ERROR_ is not > > > set to 1 for the invalid observation.
> > > They work on input statement and INPUT function but not inputN/C. As > > > far as PROC TRANSPOSE goes perhaps LET will produce the effect you > > > desire but I'm not sure exactly what you are referring to. Can you > > > make example?
> > > On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > > > Have to say, that's the first time I've come across ??. Is that the > > only > > > > place it works to suppress invalid messages, or does it work > elsewhere? > > > > (For example, a proc transpose where you have columns overwriting > other > > > > columns intentionally, etc.)?
> > > > -Joe
> > > > On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; < > > > > iebup...@gmail.com> wrote:
> > > > > It is a bit easier to quiet invalid data messages with ??
> > > > > if missing(input(var,??best12.)) then delete;
> > > > > On 1/26/09, Arthur Tabachneck <art...@netscape.net> wrote: > > > > > > Jason,
> > > > > > If Howard's concern is valid and you need to keep numbers stated > in > > > > > > scientific notation, then I'd suggest the following method:
> > > > > > >On Mon, 26 Jan 2009 11:27:55 -0800, J M <jas...@UCLA.EDU> > wrote:
> > > > > > >>How can delete observations containing characters (eg 92z89 or > > abcd)? > > > > > > >>I just want to keep observations with numeric values. > > > > > > >>Thanks.
> > > > > > >How about 12e3? It contains an alpha character but is > nevertheless > > valid > > > > > > as > > > > > > >the external representation of a numeric value?
I get it into my inbox perfectly fine :) There's big bright lines indicating his email address is not what he says it is, but it's fine. (He's not being sent to spam because of the contents of the address, but because it appears to be an email spoofing attempt; which is accurate, except for the lack of malicious intent to non-spambots.)
Filter: If From contains (Howard Schreier) Action: Never Send To Spam, Apply Label "SAS-L"
Don't create a filter based on his email address... create it based on his name. If that doesn't work, I don't know what to tell you, other than ... my gmail is nicer than yours?
- Joe
On Mon, Jan 26, 2009 at 6:38 PM, ./ ADD NAME=Data _null_; <
iebup...@gmail.com> wrote: > No Howard's e-mail refuses to be accepted by gmail as anything but > spam. I can "STAR" them and that makes them easier to find in the > spam box.
> The only think I can think of is that JUNK in the e-mail address is > too much for G-Mail's spam filter to resist.
> I continue to check my spam box throughout the day and look forward to > seeing spam that has been starred.
> Thankfully e-mail from "Ajay who's it" does end up in the spam box as > directed.
> On 1/26/09, SAS_learner <procconte...@gmail.com> wrote: > > Hello _null_,
> > Did you figure out how to get Howard's email to Gmail Inbox. If so can > you > > let me know ??
> > Thanks > > SL
> > On Mon, Jan 26, 2009 at 5:48 PM, ./ ADD NAME=Data _null_; > > <iebup...@gmail.com> wrote: > > > Format Modifiers for Error Reporting > > > ? > > > suppresses printing the invalid data note when SAS encounters invalid > > > data values.
> > > See Also: How Invalid Data is Handled
> > > ?? > > > suppresses printing the messages and the input lines when SAS > > > encounters invalid data values. The automatic variable _ERROR_ is not > > > set to 1 for the invalid observation.
> > > They work on input statement and INPUT function but not inputN/C. As > > > far as PROC TRANSPOSE goes perhaps LET will produce the effect you > > > desire but I'm not sure exactly what you are referring to. Can you > > > make example?
> > > On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > > > Have to say, that's the first time I've come across ??. Is that the > > only > > > > place it works to suppress invalid messages, or does it work > elsewhere? > > > > (For example, a proc transpose where you have columns overwriting > other > > > > columns intentionally, etc.)?
> > > > -Joe
> > > > On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; < > > > > iebup...@gmail.com> wrote:
> > > > > It is a bit easier to quiet invalid data messages with ??
> > > > > if missing(input(var,??best12.)) then delete;
> > > > > On 1/26/09, Arthur Tabachneck <art...@netscape.net> wrote: > > > > > > Jason,
> > > > > > If Howard's concern is valid and you need to keep numbers stated > in > > > > > > scientific notation, then I'd suggest the following method:
> > > > > > >On Mon, 26 Jan 2009 11:27:55 -0800, J M <jas...@UCLA.EDU> > wrote:
> > > > > > >>How can delete observations containing characters (eg 92z89 or > > abcd)? > > > > > > >>I just want to keep observations with numeric values. > > > > > > >>Thanks.
> > > > > > >How about 12e3? It contains an alpha character but is > nevertheless > > valid > > > > > > as > > > > > > >the external representation of a numeric value?
iebup...@gmail.com> wrote: > Format Modifiers for Error Reporting > ? > suppresses printing the invalid data note when SAS encounters invalid > data values.
> See Also: How Invalid Data is Handled
> ?? > suppresses printing the messages and the input lines when SAS > encounters invalid data values. The automatic variable _ERROR_ is not > set to 1 for the invalid observation.
> They work on input statement and INPUT function but not inputN/C. As > far as PROC TRANSPOSE goes perhaps LET will produce the effect you > desire but I'm not sure exactly what you are referring to. Can you > make example?
> On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > Have to say, that's the first time I've come across ??. Is that the only > > place it works to suppress invalid messages, or does it work elsewhere? > > (For example, a proc transpose where you have columns overwriting other > > columns intentionally, etc.)?
> > -Joe
> > On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; < > > iebup...@gmail.com> wrote:
> > > It is a bit easier to quiet invalid data messages with ??
> > > if missing(input(var,??best12.)) then delete;
> > > On 1/26/09, Arthur Tabachneck <art...@netscape.net> wrote: > > > > Jason,
> > > > If Howard's concern is valid and you need to keep numbers stated in > > > > scientific notation, then I'd suggest the following method:
> > > > HTH, > > > > Art > > > > ----- > > > > On Mon, 26 Jan 2009 16:01:47 -0500, Howard Schreier <hs AT dc-sug DOT > > > org> > > > > <schreier.junk.m...@GMAIL.COM> wrote:
> > > > >On Mon, 26 Jan 2009 11:27:55 -0800, J M <jas...@UCLA.EDU> wrote:
> > > > >>How can delete observations containing characters (eg 92z89 or > abcd)? > > > > >>I just want to keep observations with numeric values. > > > > >>Thanks.
> > > > >How about 12e3? It contains an alpha character but is nevertheless > valid > > > > as > > > > >the external representation of a numeric value?
> Did you figure out how to get Howard's email to Gmail Inbox. If so can you > let me know ??
> Thanks > SL
> On Mon, Jan 26, 2009 at 5:48 PM, ./ ADD NAME=Data _null_; > <iebup...@gmail.com> wrote: > > Format Modifiers for Error Reporting > > ? > > suppresses printing the invalid data note when SAS encounters invalid > > data values.
> > See Also: How Invalid Data is Handled
> > ?? > > suppresses printing the messages and the input lines when SAS > > encounters invalid data values. The automatic variable _ERROR_ is not > > set to 1 for the invalid observation.
> > They work on input statement and INPUT function but not inputN/C. As > > far as PROC TRANSPOSE goes perhaps LET will produce the effect you > > desire but I'm not sure exactly what you are referring to. Can you > > make example?
> > On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > > Have to say, that's the first time I've come across ??. Is that the > only > > > place it works to suppress invalid messages, or does it work elsewhere? > > > (For example, a proc transpose where you have columns overwriting other > > > columns intentionally, etc.)?
> > > -Joe
> > > On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; < > > > iebup...@gmail.com> wrote:
> > > > It is a bit easier to quiet invalid data messages with ??
> > > > if missing(input(var,??best12.)) then delete;
> > > > On 1/26/09, Arthur Tabachneck <art...@netscape.net> wrote: > > > > > Jason,
> > > > > If Howard's concern is valid and you need to keep numbers stated in > > > > > scientific notation, then I'd suggest the following method:
> > > > > HTH, > > > > > Art > > > > > ----- > > > > > On Mon, 26 Jan 2009 16:01:47 -0500, Howard Schreier <hs AT dc-sug > DOT > > > > org> > > > > > <schreier.junk.m...@GMAIL.COM> wrote:
> > > > > >On Mon, 26 Jan 2009 11:27:55 -0800, J M <jas...@UCLA.EDU> wrote:
> > > > > >>How can delete observations containing characters (eg 92z89 or > abcd)? > > > > > >>I just want to keep observations with numeric values. > > > > > >>Thanks.
> > > > > >How about 12e3? It contains an alpha character but is nevertheless > valid > > > > > as > > > > > >the external representation of a numeric value?
You have to create a filter specifically marking mail from him not as spam. IIRC, gmail will do that directly for you if you mark it as 'not spam' from the spam box.
_null_, the example would be something like
data test; *this is the output from a generated proc means statement; input stat $ value; datalines; T2B_q4 75.8 N_q4 80 T3B_q4 88.3 N_q4 80 ;;;; run; proc transpose data=test out =test2 let; id stat; var value; run;
The ? doesn't seem to do anything here, as far as I can tell, so that probably answers my question :) In theory I could (obviously) parse the dataset for the known duplicate values and remove them, but with the size of the data involved it's more efficient to simply ignore them, PROC PRINTTO to send my log to a dummy file, and not worry about it (as much as I hate leaving that bit of the log out). Otherwise with the number of observations I have, it takes a dozen or so clears of the log while it's running, and takes longer to write the log out than to do the proc :)
On Mon, Jan 26, 2009 at 6:11 PM, SAS_learner <procconte...@gmail.com> wrote: > Hello _null_,
> Did you figure out how to get Howard's email to Gmail Inbox. If so can you > let me know ??
> Thanks > SL
> On Mon, Jan 26, 2009 at 5:48 PM, ./ ADD NAME=Data _null_; < > iebup...@gmail.com> wrote:
> > Format Modifiers for Error Reporting > > ? > > suppresses printing the invalid data note when SAS encounters invalid > > data values.
> > See Also: How Invalid Data is Handled
> > ?? > > suppresses printing the messages and the input lines when SAS > > encounters invalid data values. The automatic variable _ERROR_ is not > > set to 1 for the invalid observation.
> > They work on input statement and INPUT function but not inputN/C. As > > far as PROC TRANSPOSE goes perhaps LET will produce the effect you > > desire but I'm not sure exactly what you are referring to. Can you > > make example?
> > On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > > Have to say, that's the first time I've come across ??. Is that the > only > > > place it works to suppress invalid messages, or does it work elsewhere? > > > (For example, a proc transpose where you have columns overwriting other > > > columns intentionally, etc.)?
> > > -Joe
> > > On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; < > > > iebup...@gmail.com> wrote:
> > > > It is a bit easier to quiet invalid data messages with ??
> > > > if missing(input(var,??best12.)) then delete;
> > > > On 1/26/09, Arthur Tabachneck <art...@netscape.net> wrote: > > > > > Jason,
> > > > > If Howard's concern is valid and you need to keep numbers stated in > > > > > scientific notation, then I'd suggest the following method:
> > > > > HTH, > > > > > Art > > > > > ----- > > > > > On Mon, 26 Jan 2009 16:01:47 -0500, Howard Schreier <hs AT dc-sug > DOT > > > > org> > > > > > <schreier.junk.m...@GMAIL.COM> wrote:
> > > > > >On Mon, 26 Jan 2009 11:27:55 -0800, J M <jas...@UCLA.EDU> wrote:
> > > > > >>How can delete observations containing characters (eg 92z89 or > > abcd)? > > > > > >>I just want to keep observations with numeric values. > > > > > >>Thanks.
> > > > > >How about 12e3? It contains an alpha character but is nevertheless > > valid > > > > > as > > > > > >the external representation of a numeric value?
No idea... these are the options I get when I create a filter, and then "Next Step": (They're images, so I imagine this won't look interesting on the actual list - sorry!)
Choose action - Now, select the action you'd like to take on messages that match the criteria you specified.When a message arrives that matches the search: from:(Howard Schreier), do the following: Skip the Inbox (Archive it) Mark as read Star it Apply the label: New label...-----RestSASLuseful sas Forward it to: Delete it Never send it to Spam Show current filtersCancel=AB BackUpdate Filter Also apply filter to 170 conversations below.
Text: * Skip the Inbox (Archive It) * Mark as read * Star it * Apply the label: (labels) * Forward it to: (textbox) * Delete it * Never send it to Spam
I got to that from Settings -> Filters -> edit the particular filter -> Nex= t Step.
-Joe
On Mon, Jan 26, 2009 at 7:22 PM, ./ ADD NAME=3DData _null_; <
iebup...@gmail.com> wrote: > Never Send To Spam is not a choice when I create a filter. What is > different? I thought gmail was gmail.
> On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > I get it into my inbox perfectly fine :) There's big bright lines > > indicating his email address is not what he says it is, but it's fine. > > (He's not being sent to spam because of the contents of the address, bu= t > > because it appears to be an email spoofing attempt; which is accurate, > > except for the lack of malicious intent to non-spambots.)
> > Filter: If From contains (Howard Schreier) > > Action: Never Send To Spam, Apply Label "SAS-L"
> > Don't create a filter based on his email address... create it based on > his > > name. If that doesn't work, I don't know what to tell you, other than > ... > > my gmail is nicer than yours?
> > - > > Joe
> > On Mon, Jan 26, 2009 at 6:38 PM, ./ ADD NAME=3DData _null_; > > <iebup...@gmail.com> wrote: > > > No Howard's e-mail refuses to be accepted by gmail as anything but > > > spam. I can "STAR" them and that makes them easier to find in the > > > spam box.
> > > The only think I can think of is that JUNK in the e-mail address is > > > too much for G-Mail's spam filter to resist.
> > > I continue to check my spam box throughout the day and look forward t= o > > > seeing spam that has been starred.
> > > Thankfully e-mail from "Ajay who's it" does end up in the spam box as > > directed.
> > > > Did you figure out how to get Howard's email to Gmail Inbox. If so > can > > you > > > > let me know ??
> > > > Thanks > > > > SL
> > > > On Mon, Jan 26, 2009 at 5:48 PM, ./ ADD NAME=3DData _null_; > > > > <iebup...@gmail.com> wrote: > > > > > Format Modifiers for Error Reporting > > > > > ? > > > > > suppresses printing the invalid data note when SAS encounters > invalid > > > > > data values.
> > > > > See Also: How Invalid Data is Handled
> > > > > ?? > > > > > suppresses printing the messages and the input lines when SAS > > > > > encounters invalid data values. The automatic variable _ERROR_ is > not > > > > > set to 1 for the invalid observation.
> > > > > They work on input statement and INPUT function but not inputN/C. > As > > > > > far as PROC TRANSPOSE goes perhaps LET will produce the effect yo= u > > > > > desire but I'm not sure exactly what you are referring to. Can y= ou > > > > > make example?
> > > > > On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > > > > > Have to say, that's the first time I've come across ??. Is tha= t > the > > > > only > > > > > > place it works to suppress invalid messages, or does it work > > elsewhere? > > > > > > (For example, a proc transpose where you have columns overwriti= ng > > other > > > > > > columns intentionally, etc.)?
> > > > > > -Joe
> > > > > > On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=3DData _null_; < > > > > > > iebup...@gmail.com> wrote:
> > > > > > > It is a bit easier to quiet invalid data messages with ??
> > > > > > > if missing(input(var,??best12.)) then delete;
> > > > > > > > If Howard's concern is valid and you need to keep numbers > stated > > in > > > > > > > > scientific notation, then I'd suggest the following method:
> I get it into my inbox perfectly fine :) There's big bright lines > indicating his email address is not what he says it is, but it's fine. > (He's not being sent to spam because of the contents of the address, but > because it appears to be an email spoofing attempt; which is accurate, > except for the lack of malicious intent to non-spambots.)
> Filter: If From contains (Howard Schreier) > Action: Never Send To Spam, Apply Label "SAS-L"
> Don't create a filter based on his email address... create it based on his > name. If that doesn't work, I don't know what to tell you, other than ... > my gmail is nicer than yours?
> - > Joe
> On Mon, Jan 26, 2009 at 6:38 PM, ./ ADD NAME=Data _null_; > <iebup...@gmail.com> wrote: > > No Howard's e-mail refuses to be accepted by gmail as anything but > > spam. I can "STAR" them and that makes them easier to find in the > > spam box.
> > The only think I can think of is that JUNK in the e-mail address is > > too much for G-Mail's spam filter to resist.
> > I continue to check my spam box throughout the day and look forward to > > seeing spam that has been starred.
> > Thankfully e-mail from "Ajay who's it" does end up in the spam box as > directed.
> > > Did you figure out how to get Howard's email to Gmail Inbox. If so can > you > > > let me know ??
> > > Thanks > > > SL
> > > On Mon, Jan 26, 2009 at 5:48 PM, ./ ADD NAME=Data _null_; > > > <iebup...@gmail.com> wrote: > > > > Format Modifiers for Error Reporting > > > > ? > > > > suppresses printing the invalid data note when SAS encounters invalid > > > > data values.
> > > > See Also: How Invalid Data is Handled
> > > > ?? > > > > suppresses printing the messages and the input lines when SAS > > > > encounters invalid data values. The automatic variable _ERROR_ is not > > > > set to 1 for the invalid observation.
> > > > They work on input statement and INPUT function but not inputN/C. As > > > > far as PROC TRANSPOSE goes perhaps LET will produce the effect you > > > > desire but I'm not sure exactly what you are referring to. Can you > > > > make example?
> > > > On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > > > > Have to say, that's the first time I've come across ??. Is that the > > > only > > > > > place it works to suppress invalid messages, or does it work > elsewhere? > > > > > (For example, a proc transpose where you have columns overwriting > other > > > > > columns intentionally, etc.)?
> > > > > -Joe
> > > > > On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; < > > > > > iebup...@gmail.com> wrote:
> > > > > > It is a bit easier to quiet invalid data messages with ??
> > > > > > if missing(input(var,??best12.)) then delete;
> > > > > > On 1/26/09, Arthur Tabachneck <art...@netscape.net> wrote: > > > > > > > Jason,
> > > > > > > If Howard's concern is valid and you need to keep numbers stated > in > > > > > > > scientific notation, then I'd suggest the following method:
> > > > > > > >On Mon, 26 Jan 2009 11:27:55 -0800, J M <jas...@UCLA.EDU> > wrote:
> > > > > > > >>How can delete observations containing characters (eg 92z89 or > > > abcd)? > > > > > > > >>I just want to keep observations with numeric values. > > > > > > > >>Thanks.
> > > > > > > >How about 12e3? It contains an alpha character but is > nevertheless > > > valid > > > > > > > as > > > > > > > >the external representation of a numeric value?
I did create a filter it does everything but put Howard's e-mail in the right box. You use gmail have you been successful. Maybe it has to do with versions of IE or something.
As for the Transpose I would take a step back to create the output from PROC MEANS that was better suited to my needs. If you want to talk about that it would make an interesting, I think, topic for a new thread.
On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote:
> You have to create a filter specifically marking mail from him not as spam. > IIRC, gmail will do that directly for you if you mark it as 'not spam' from > the spam box.
> _null_, the example would be something like
> data test; *this is the output from a generated proc means statement; > input stat $ value; > datalines; > T2B_q4 75.8 > N_q4 80 > T3B_q4 88.3 > N_q4 80 > ;;;; > run; > proc transpose data=test out =test2 let; > id stat; > var value; > run;
> The ? doesn't seem to do anything here, as far as I can tell, so that > probably answers my question :) In theory I could (obviously) parse the > dataset for the known duplicate values and remove them, but with the size of > the data involved it's more efficient to simply ignore them, PROC PRINTTO to > send my log to a dummy file, and not worry about it (as much as I hate > leaving that bit of the log out). Otherwise with the number of observations > I have, it takes a dozen or so clears of the log while it's running, and > takes longer to write the log out than to do the proc :)
> -Joe
> On Mon, Jan 26, 2009 at 6:11 PM, SAS_learner <procconte...@gmail.com> wrote:
> > Hello _null_,
> > Did you figure out how to get Howard's email to Gmail Inbox. If so can you > > let me know ??
> > Thanks > > SL
> > On Mon, Jan 26, 2009 at 5:48 PM, ./ ADD NAME=Data _null_; < > > iebup...@gmail.com> wrote:
> > > Format Modifiers for Error Reporting > > > ? > > > suppresses printing the invalid data note when SAS encounters invalid > > > data values.
> > > See Also: How Invalid Data is Handled
> > > ?? > > > suppresses printing the messages and the input lines when SAS > > > encounters invalid data values. The automatic variable _ERROR_ is not > > > set to 1 for the invalid observation.
> > > They work on input statement and INPUT function but not inputN/C. As > > > far as PROC TRANSPOSE goes perhaps LET will produce the effect you > > > desire but I'm not sure exactly what you are referring to. Can you > > > make example?
> > > On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > > > Have to say, that's the first time I've come across ??. Is that the > > only > > > > place it works to suppress invalid messages, or does it work elsewhere? > > > > (For example, a proc transpose where you have columns overwriting other > > > > columns intentionally, etc.)?
> > > > -Joe
> > > > On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=Data _null_; < > > > > iebup...@gmail.com> wrote:
> > > > > It is a bit easier to quiet invalid data messages with ??
> > > > > if missing(input(var,??best12.)) then delete;
> > > > > On 1/26/09, Arthur Tabachneck <art...@netscape.net> wrote: > > > > > > Jason,
> > > > > > If Howard's concern is valid and you need to keep numbers stated in > > > > > > scientific notation, then I'd suggest the following method:
> > > > > > >On Mon, 26 Jan 2009 11:27:55 -0800, J M <jas...@UCLA.EDU> wrote:
> > > > > > >>How can delete observations containing characters (eg 92z89 or > > > abcd)? > > > > > > >>I just want to keep observations with numeric values. > > > > > > >>Thanks.
> > > > > > >How about 12e3? It contains an alpha character but is nevertheless > > > valid > > > > > > as > > > > > > >the external representation of a numeric value?
On Mon, Jan 26, 2009 at 8:33 PM, Joe Matise <snoopy...@gmail.com> wrote: > No idea... these are the options I get when I create a filter, and then > "Next Step": (They're images, so I imagine this won't look interesting on > the actual list - sorry!)
> Choose action - Now, select the action you'd like to take on messages tha= t > match the criteria you specified.When a message arrives that matches the > search: from:(Howard Schreier), do the following: Skip the Inbox (Archiv= e > it) > Mark as read > Star it > Apply the label: New label...-----RestSASLuseful sas > Forward it to: > Delete it > Never send it to Spam > Show current filtersCancel=AB BackUpdate Filter Also apply filter to 17= 0 > conversations below.
> Text: > * Skip the Inbox (Archive It) > * Mark as read > * Star it > * Apply the label: (labels) > * Forward it to: (textbox) > * Delete it > * Never send it to Spam
> I got to that from Settings -> Filters -> edit the particular filter -> > Next > Step.
> -Joe
> On Mon, Jan 26, 2009 at 7:22 PM, ./ ADD NAME=3DData _null_; < > iebup...@gmail.com> wrote:
> > Never Send To Spam is not a choice when I create a filter. What is > > different? I thought gmail was gmail.
> > On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > > I get it into my inbox perfectly fine :) There's big bright lines > > > indicating his email address is not what he says it is, but it's fine= . > > > (He's not being sent to spam because of the contents of the address, > but > > > because it appears to be an email spoofing attempt; which is accurate= , > > > except for the lack of malicious intent to non-spambots.)
> > > Filter: If From contains (Howard Schreier) > > > Action: Never Send To Spam, Apply Label "SAS-L"
> > > Don't create a filter based on his email address... create it based o= n > > his > > > name. If that doesn't work, I don't know what to tell you, other tha= n > > ... > > > my gmail is nicer than yours?
> > > - > > > Joe
> > > On Mon, Jan 26, 2009 at 6:38 PM, ./ ADD NAME=3DData _null_; > > > <iebup...@gmail.com> wrote: > > > > No Howard's e-mail refuses to be accepted by gmail as anything but > > > > spam. I can "STAR" them and that makes them easier to find in the > > > > spam box.
> > > > The only think I can think of is that JUNK in the e-mail address is > > > > too much for G-Mail's spam filter to resist.
> > > > I continue to check my spam box throughout the day and look forward > to > > > > seeing spam that has been starred.
> > > > Thankfully e-mail from "Ajay who's it" does end up in the spam box = as > > > directed.
> > > > > Did you figure out how to get Howard's email to Gmail Inbox. If s= o > > can > > > you > > > > > let me know ??
> > > > > Thanks > > > > > SL
> > > > > On Mon, Jan 26, 2009 at 5:48 PM, ./ ADD NAME=3DData _null_; > > > > > <iebup...@gmail.com> wrote: > > > > > > Format Modifiers for Error Reporting > > > > > > ? > > > > > > suppresses printing the invalid data note when SAS encounters > > invalid > > > > > > data values.
> > > > > > See Also: How Invalid Data is Handled
> > > > > > ?? > > > > > > suppresses printing the messages and the input lines when SAS > > > > > > encounters invalid data values. The automatic variable _ERROR_ = is > > not > > > > > > set to 1 for the invalid observation.
> > > > > > They work on input statement and INPUT function but not inputN/= C. > > As > > > > > > far as PROC TRANSPOSE goes perhaps LET will produce the effect > you > > > > > > desire but I'm not sure exactly what you are referring to. Can > you > > > > > > make example?
> > > > > > On 1/26/09, Joe Matise <snoopy...@gmail.com> wrote: > > > > > > > Have to say, that's the first time I've come across ??. Is > that > > the > > > > > only > > > > > > > place it works to suppress invalid messages, or does it work > > > elsewhere? > > > > > > > (For example, a proc transpose where you have columns > overwriting > > > other > > > > > > > columns intentionally, etc.)?
> > > > > > > -Joe
> > > > > > > On Mon, Jan 26, 2009 at 3:42 PM, ./ ADD NAME=3DData _null_; < > > > > > > > iebup...@gmail.com> wrote:
> > > > > > > > It is a bit easier to quiet invalid data messages with ??
> > > > > > > > if missing(input(var,??best12.)) then delete;
> > > > > > > > > If Howard's concern is valid and you need to keep numbers > > stated > > > in > > > > > > > > > scientific notation, then I'd suggest the following metho= d:
|
