cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 
sastpw
‎12-24-2025
  • 203 Posts
  • 39 Likes Given
  • 28 Solutions
  • 77 Likes Received

Re: SASPy - How to automate the population of SAS Config, user ID, and...

by SAS Employee in SAS Programming
‎11-26-2019 09:55 AM
‎11-26-2019 09:55 AM
Hey, I'm a little confused. saspy has a way for you to store credentials in a secure file so that you don't need to do any of the things you're trying to do, based upon the other users suggestion, which was basically, try to implement that yourself. That was a fine idea, except there's no need, and trying to implement it yourself doesn't get you working when trying to use the magic.   Did you see my post about the authinfo file, with the instructions and the link to the documentation? If you put your credentials in the authinfo file, then saspy will just work without prompting you for all cases. And you don't need to write special code to try to implement it yourself.   Two other things worth noting. the magic creates its own SASsession to use. That was the default behavior when it was originally implemented. Since then it was enhanced to take a SASsession object and use that session to submit the code.The magic just calls the submit() method and displays the LST, if there is one, else the LOG.    In the current release, V3.1.8, there are two new variants of the submit method; submitLOG() and submitLST() (these are in the API doc), They are like the magic in that instead of returning the LOG and LST to you in a dictionary (as submit does), instead the submitLOG just displays the LOG and the submitLST displays the LST (results). submitLST also has a method= paramter to allow you to have it display the log if there is no LST, or to display both in either order you like. The default is the LST whether there is any or not.   So, sas.submitLST("sas code", method='listorlog') is basically the same as the magic. If there output it's display that directly. If there's no output, the LOG is displayed instead.   Either of these last two things will get you a work around the issue with not being able use the magic with your self implemented credentials, but I would suggest using the supported method instead, and not having any of these issues to begin with.   Tom ... View more

Re: SASPy - How to automate the population of SAS Config, user ID, and...

by SAS Employee in SAS Programming
‎11-25-2019 09:05 AM
‎11-25-2019 09:05 AM
If you used the authinfo file for your credentials, that saspy supports, this would work too. Tom ... View more

Re: SASPy - How to automate the population of SAS Config, user ID, and...

by SAS Employee in SAS Programming
‎11-21-2019 03:33 PM
‎11-21-2019 03:33 PM
Hey, yes, you can simply put your credentials in an authinfo file and then specify the key for that in your config. See the doc here: https://sassoftware.github.io/saspy/install.html#iom-using-java and scroll down to the authinfo part. Well, I'll cut-n-paste that next. You use the 'authkey' in your configuration definition to specify which credentials to use. You can have different credentials for different connections.       The IOM access method now has support for getting the required user/password from an authinfo file in the user’s home directory instead of prompting for it. On linux, the file is named .authinfo and on windows, it’s _authinfo. The format of the line in the authinfo file is as follows. The first value is the authkey value you specify for  authkey . Next is the ‘user’ key followed by the value (the user id) and then ‘password’ key followed by its value (the user’s password). Note that there are permission rules for this file. On linux the file must have permissions of 600, only the user can read or write the file. On Windows, the file should be equally locked down to where only the owner can read and write it. authkey user omr_user_id password omr_user_password So, for a Configuration Definition that specifies the following authkey: 'authkey' : 'IOM_Prod_Grid1', The authinfo file in the home directory for user Bob, with a password of BobsPW1 would have a line in it as follows: IOM_Prod_Grid1 user Bob password BobsPW1   Thanks, Tom   ... View more

Re: SAS over SSH (with saspy)

by SAS Employee in New SAS User
‎10-18-2019 01:20 PM
1 Like
‎10-18-2019 01:20 PM
1 Like
I wanted to try this out myself on 9.2, so I could diagnose what's happening, at least. But, I can't get at 9.2. In our dev environment I can run various releases, but it no longer goes back past 9.30. I'd have to find a real install or something not readily obvious.If I were at the keyboard where this was running, I'm sure I could identify what the problems are for things that don't work, and maybe figure a way around them, and tell what would work. But at this point I'm just speculating. As @Tom mentioned, you have 9.4 available from us for free, as you're a customer, so getting a current version of SAS would certainly make things easier.   Let me know if there's anything else I can do.   Thanks, Tom ... View more

Re: SAS over SSH (with saspy)

by SAS Employee in New SAS User
‎10-18-2019 10:57 AM
1 Like
‎10-18-2019 10:57 AM
1 Like
Ok, so that confirms my suspicion that it's something more general than proc export. This won't work completely with 9.2. but, using the STDIO, over SSH or not, will work in many ways even with older versions of SAS. I can only guess at what the exact issue happens to be for specific case. That's where being able to try things would help. Using the sas.saslog() method would be helpful to see what's happening. Setting results='text' on the SASsession object can maybe help for some things. There's just no guessing at what may trip up saspy for any given thing you submit since it's not made to run against that old of a version. So there will be things that work and things that don't. Tom ... View more

Re: SAS over SSH (with saspy)

by SAS Employee in New SAS User
‎10-18-2019 09:47 AM
‎10-18-2019 09:47 AM
So that shows that the export from SAS dataset to cvs, on the SAS server isn't happening, of finishing or something? That's the first step. That seems very strange.There's more going on here than the usual situation.I need more information to diagnose what's going on here. I often got on a web-ex (I can set that up) with customers when trying to post back and forth doesn't seem to get us to the answer very quickly. Is that something you could do? I could diagnose this much quicker interactively seeing what's going on. If not, we can continue this way. The first thing I might have you try is to set obs=10 so it can't possibly be a size issue and see if it still hangs or whatever it's doing.  I don't see any output from anything, so there's not much to go on. dsopts={'keep' : 'column', 'obs': 10}) I'd like to see a number of other things to, but that's where it would be much more efficient to look at this interactively if that's an option. I'm not sure if it's really the export to csv that's the issue, rather than something more general.    Also, I would probably open in issue on saspy for this (https://github.com/sassoftware/saspy/issues), though as I'm already on this here, that can go either way.   Thanks, Tom ... View more

Re: SAS over SSH (with saspy)

by SAS Employee in New SAS User
‎10-17-2019 10:59 AM
‎10-17-2019 10:59 AM
The other simple thing is that if you are using sd2df(), you really should be using sd2df_CSV() since it performs better with larger data. BUT, I'd like to see the manual steps in the previous post so we can determine where the problem is. Again, sd2df_CSV() does all these things itself, but I'd like to do each step manually to diagnose what it really the problem.   Thanks! Tom ... View more

Re: SAS over SSH (with saspy)

by SAS Employee in New SAS User
‎10-17-2019 10:55 AM
‎10-17-2019 10:55 AM
Ok, so, what code exactly have you run? I showed 3 different possibilities earlier, It sounds like one of two things (or maybe part of the same). That the data you're trying to load into python is more than that process can support. Or, it's so large and you're using the sd2df() method and it's taking so long that it's not workable.   Can you try the following and see what happens:   table = sas.sasdata('your_table','libref_if_not_work', dsopts={'keep' : 'column1 column2'})   table.to_csv(sas.workpath+'filename.csv')    sas.download('local_file_path', sas.workpath+'filename.csv')    import pandas df = pandas.read_csv(('local_file_path/'+'filename.csv)   I expect that the first two; creating the csv and downloading it won't take terribly long. And I expect the pandas.read_csv() will show if it's really too much data for your python process or if can work. The above is basically doing what the following does, but in individual steps so you can see if there's a problem, and with which one. df = table.to_df_CSV(tempfile='local_file_path/filename.csv, tempkeep=True)     Thanks, Tom ... View more

Re: SAS over SSH (with saspy)

by SAS Employee in New SAS User
‎10-17-2019 09:32 AM
‎10-17-2019 09:32 AM
I'm not sure I completely understand, so I can't give you the exact answer yet. You are using SSH, so the following is what I believe is true:   1) SAS is on a remote system 2) python, and thus saspy, are on your local system 3) the large SAS data set is on the remote SAS server 4) you want 2 columns of that large dataset where? In python as a pandas data frame (that's on your client)? on your client system as a .csv file? Somewhere else as a csv?   Can you explain exactly what you want where and I can provide a better answer.    Thanks! Tom ... View more

Re: SAS over SSH (with saspy)

by SAS Employee in New SAS User
‎10-16-2019 12:03 PM
1 Like
‎10-16-2019 12:03 PM
1 Like
Hey, to do this with saspy, is fairly simple. Not knowing your talbe or anything, I'll make up names, but you would change them for your data.   import saspy sas = saspy.SASsession() sas   table = sas.sasdata('your_table','libref_if_not_work', dsopts={'keep' : 'column1 column2'}) df = table.to_df()  #now you have a dataframe with the two columns (col1 and col2) from the table in the libref you wanted.   As for saving to csv, you can do that from the dataframe (df.to_csv()  that's the pandas method). Or, you could do that from SAS without bringing the data over to python inthe first place:   table.to_csv('path/filename.csv')  # now you have a .csv file with the two columns. this is the saspy to_csv() method   If you are running a local SAS, the csv file is on your machine. If it's a remote (which you said SSH, so it's remote), you can use the download method to pull it over to your local machine, if you need it there.   sas.download('local_file_path', 'path/filename.csv')  # now you have local_file_path/filename.csv file on your machine     There's even a way to create the dataframe and create the csv file and have it be copied to your local machine all in one line of code, if that's what you want; sort of all of the above in one:   df = table.to_df_CSV(tempfile='local_file_path/filename.csv, tempkeep=True)  # now you have data frame and local .csv file too   So, you see, with only a line or two of python code, you can accomplish this a variety ow ways, depending upon your needs.   Let me know if any of these work for you or if there's something slightly different you need? Might just be another line of code.   Thanks, Tom ... View more

Re: SASPy - Reference existing table for filtering on proc sql stateme...

by SAS Employee in SAS Programming
‎09-20-2019 09:31 AM
‎09-20-2019 09:31 AM
Yep, that or the join. If you need any other help with this, just let me know!   Thanks, Tom ... View more

Re: SASPy - Reference existing table for filtering on proc sql stateme...

by SAS Employee in SAS Programming
‎09-20-2019 07:51 AM
‎09-20-2019 07:51 AM
Ok, well, Teradata (ans SAS) support the in() operator, but if you have a list that's so long it's hitting some limit, then that (even if it didn't hit the limit), would be a very inefficient way to do a query. I would create a temp table with the values and use the subquery instead, or even better, create the temp table with distinct values and do an inner join with it in your query. That will perform the best.   Tom ... View more

Re: SASPy - Reference existing table for filtering on proc sql stateme...

by SAS Employee in SAS Programming
‎09-19-2019 04:41 PM
‎09-19-2019 04:41 PM
Hey, just off the top of my head; I haven't tried it.Could you use a subquery instead of a list. create a table with one column and all of those values as the rows, then have your where clause criteria be the subquery instead of the list: ``` where a.member in (select mem from temp_table) ```    As I don't see any real saspy code, this is just SAS proc sql, I'm guessing the list of members came from python? You can have that as a dataframe and then use df2sd() to create the table (even in teradata) to then use it in the sql.   If I see more code (saspy, python side), I could probably code up something more specific.   Tom ... View more

Re: Passing credentials in saslib method

by SAS Employee in Developers
‎08-28-2019 10:24 AM
‎08-28-2019 10:24 AM
Hey, I think you can just quote the pw and it ought to work. Give this a try:   sas.saslib('rpt', engine='Teradata', options='user='+username+' pw="'+password+'" server=server1 schema=schema1')   Hopefully that's all you need. Tom  ... View more

Re: Sas Authkey

by SAS Employee in SAS Programming
‎08-19-2019 09:05 AM
‎08-19-2019 09:05 AM
Great, I'm glad that's all it was. BTW, there's another configuration definition key you can add to your configuration definition to has the validvarname option set every time you connect (if that's what you want). There's an 'autoexec':'string of SAS code;' which will be automatically executed for you when a session is established. So if you want validvarname=any every time you connect using a given configuration definition you can add 'autoexec' : 'options validvarname=any;' and it will automatically be set for you.   From the doc: autoexec -This is a string of SAS code that will be submitted upon establishing a connection. You can use this to preassign libraries you always want available, or whatever you want. Don’t confuse this with the autoexec option of SAS which specifies a sas program file to be run. That is different. This is a string of SAS code saspy will submit after the session is created, which would be after SAS already included any autoexec file if there was one.   Tom ... View more
Contact Me
Online Status
Offline
Date Last Visited
‎12-24-2025 09:01 AM
Group Hubs for sastpw
View All
Likes given to
Likes from
Quick links