SAS Vs SPSS

[Krishna Chaitanya-7]

Hi SPSSers,

 

I am Krishna working as a consultant. I am comparing SAS server package with the SPSS Server for one of my clients. I have some basic doubts on how SPSS fares with SAS on the below topics --

4  Ability to extract data from an Oracle/ACCESS database

4  Handling huge amounts of data (Is there an alternate to proc sql and proc transpose)

4  Processing time

4  Opening multiple sessions (Will 15 people be able to connect simultaneously?)

4  Remote connection (How fast is this?)

4  Cost – Long term and short term

 

Any help in this topic is highly appreciated.

 

Thanks,
Krishna

 

***********************************************************************
This e-mail and any files transmitted with it are for the sole use of the intended recipient(s) and may contain confidential and privileged information. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies and the original message. Any unauthorized review, use, disclosure, dissemination, forwarding, printing or copying of this email or any action taken in reliance on this e-mail is strictly prohibited and may be unlawful. 

 The recipient acknowledges that Mu Sigma or its subsidiaries and associated companies (collectively " Mu Sigma Inc), are unable to exercise control or ensure or guarantee the integrity of over the contents of the information contained in e-mail transmissions and further acknowledges that any views expressed in this message are those of the individual sender and no binding nature of the message shall be implied or assumed unless the sender does so expressly with due authority of Mu Sigma. Before opening any attachments please check them for viruses and defects.
 ***********************************************************************


	
	
	
	

[ajayohri]

I wrote an article on SAS versus SPSS some time back. Most of it is relevant.

http://www.decisionstats.com/2008/02/comparing-sas-and-spss/

Having used SAS on both a Linux server and using SAS/Connect on remote db's I was quite happy as an analyst with the speed at which it handled data particularly sorting and merging.

Processing time is actually a function of CPUs processors and server disk capacity- SPSS lags slightly but not too much in this.

Remote connection is similarly more a function of Windows Terminal server and /or Linux Server OS, Bandwidth , and the server capacity. Generally I have seen servers to be underpowered as no one expects peak capacity say 15 people at the same time.

SAS is premium end and SPSS is middle end of analytics market so price comparisons are similarly high and medium. Answer is how mission critical is your work to be done-- people compromise on software costs upfront and later huge costs pile up in analysts waiting time or delayed decisions. 

Since Mu sigma is an outsourcing company sometimes this can affect delivery.

There are interesting variations to SAS Server Instals if it is data manipulation that you need more ( modeling generally requires smaller data). You can contact Phil Rack at Minequest ( he resells a Base SAS equivalent called WPS that can run on servers and costs are one tenth and speed is 120 % faster allegedly.

Phil Rack also has a software that can run parallel sessions of R at 50 $ a license .Its called WPS 2 R and it helps WPS with advanced technologies.

if you need to build models but score them on a huge data ( see Zementis /Mike zelleR)

Or you can contact the WPS company directly at http://teamwpc.co.uk/ 

if it is an India purchase.( see Martin Jupp cced here)

All these options have various price points - and SPSS and WPS do offer evaluation version to their credit so you can pick and choose.

Best Regards,

Ajay Ohri

www.decisionstats.com

ps Give my regards to the Dhiraj Rajaram , he is a nice guy and ask him about the strike price .

On Mon, Mar 9, 2009 at 9:35 AM, Krishna Chaitanya <[hidden email]> wrote:

Hi SPSSers,

 

I am Krishna working as a consultant. I am comparing SAS server package with the SPSS Server for one of my clients. I have some basic doubts on how SPSS fares with SAS on the below topics --

4  Ability to extract data from an Oracle/ACCESS database

4  Handling huge amounts of data (Is there an alternate to proc sql and proc transpose)

4  Processing time

4  Opening multiple sessions (Will 15 people be able to connect simultaneously?)

4  Remote connection (How fast is this?)

4  Cost – Long term and short term

 

Any help in this topic is highly appreciated.

 

Thanks,
Krishna

 

***********************************************************************
This e-mail and any files transmitted with it are for the sole use of the intended recipient(s) and may contain confidential and privileged information. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies and the original message. Any unauthorized review, use, disclosure, dissemination, forwarding, printing or copying of this email or any action taken in reliance on this e-mail is strictly prohibited and may be unlawful. 

 The recipient acknowledges that Mu Sigma or its subsidiaries and associated companies (collectively " Mu Sigma Inc), are unable to exercise control or ensure or guarantee the integrity of over the contents of the information contained in e-mail transmissions and further acknowledges that any views expressed in this message are those of the individual sender and no binding nature of the message shall be implied or assumed unless the sender does so expressly with due authority of Mu Sigma. Before opening any attachments please check them for viruses and defects.
 ***********************************************************************

[Art Kendall]

A lot depends on the particular situation.
SQL etc is external to either package. speed of access, number of users, remote connection. is very much more  a function of your network and platform not the stat software.

For large files where only one pass is required SPSS used to have a big advantage because it only created the scratch files when thy were going to be reread.  I do not know if SAS has caught up on this or not.

A major consideration is the total cost of ownership, purchase, user salary, machine resources, etc.

Purchase and license costs are often small relative to hardware, networking etc. Combined the purchase and platform costs are often small relative to people costs.

In contexts where people and calendar time are a large part of the consideration, SPSS has long had an advantage.  SPSS syntax is  more user friendly.  It is very consistent in its usage.  This tend to result in less time to write the syntax, "reference it" (audit-speak for the process of another person going over the syntax to see if it appears to do what it is supposed to), refine/maintain it, communicate to help sources, etc.  The syntax advantages are mostly in the part of the process that takes up most of the time on a project, assembling, cleaning, and other preparation of the data for analysis.

Of course, the portion of total cost that is for people, will be less of a consideration in some environments where a process is set up and then run as is for a long time doing a few things.  People costs would tend to be larger when there are many different ad hoc things that need to be done by a wide variety of people and smaller when a few people people do pretty much the same thing.


If you have the opportunity get syntax to do data cleaning, preparation, etc. tasks, i.e.,  a lot of data transformations, recodes, autorecodes, scoring, right/wrong marking, etc from both both packages.
Show both to some people who use neither package.  See how well they can tell what the syntax is doing.

By way of disclosure my experience is mostly on the kind of applications where researchers/scientist/faculty/analysts  use stat software intermittently.  Much less of it is on the kind of things like payroll, inventory, etc.  where a few people do a small range of things.

HTH
Art Kendall
Social Research Consultants

Krishna Chaitanya wrote:

Hi SPSSers,

 

I am Krishna working as a consultant. I am comparing SAS server package with the SPSS Server for one of my clients. I have some basic doubts on how SPSS fares with SAS on the below topics --

4  Ability to extract data from an Oracle/ACCESS database

4  Handling huge amounts of data (Is there an alternate to proc sql and proc transpose)

4  Processing time

4  Opening multiple sessions (Will 15 people be able to connect simultaneously?)

4  Remote connection (How fast is this?)

4  Cost – Long term and short term

 

Any help in this topic is highly appreciated.

 

Thanks,
Krishna

 

***********************************************************************
This e-mail and any files transmitted with it are for the sole use of the intended recipient(s) and may contain confidential and privileged information. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies and the original message. Any unauthorized review, use, disclosure, dissemination, forwarding, printing or copying of this email or any action taken in reliance on this e-mail is strictly prohibited and may be unlawful.

 The recipient acknowledges that Mu Sigma or its subsidiaries and associated companies (collectively " Mu Sigma Inc), are unable to exercise control or ensure or guarantee the integrity of over the contents of the information contained in e-mail transmissions and further acknowledges that any views expressed in this message are those of the individual sender and no binding nature of the message shall be implied or assumed unless the sender does so expressly with due authority of Mu Sigma. Before opening any attachments please check them for viruses and defects.
 ***********************************************************************


  

[Weeks, Kyle]

Hello Krishna,

 

As a follow up to Art Kendall’s informative response, I wanted to add a few items that will of interest to you in your evaluation.

 

SPSS provide high performance drivers for Oracle data extraction and Microsoft makes a driver available for ACESS (for free). This will enable extraction of data from both the Oracle and ACCESS database systems into SPSS Statistics Server.

 

SPSS Statistics Server can process and manipulate large amounts of data. While there may be a perception is that the product is a tool suited to smaller samples and analysis limited to survey research work, it is in fact a very robust system that can conduct complex data preparation and analysis of large volumes of data. Included with SPSS Statistics Server is SPSS Statistics Batch Facility that allows for command line batch processing.  Statistics Batch Facility is often utilized by customers to conduct “lights out” data preparation and analysis tasks and even report creation. This has the advantage of allowing the processes to take place overnight, or on the weekend, when client connections and additional activity is at a minimum.  A valuable source of information regarding the applicability of the SPSS Statistics technology to large data and complex preparation and analysis is the “SPSS Programming and Data Management” book (Fourth Edition).

 

Finally, the ability to handle multiple connections is dependent on hardware in terms of performance. 15 people can connect simultaneously with ease, but speed at which SPSS Statistics Server can accommodate process requests from each user depends on the hardware, as is the case with any software. A white paper detailing hardware requirements based on many factors, including the number of client connections, can be found at http://www.spss.com/statistics/ under the “Enterprise” tab titled “SPSS Statistics Server: New hardware recommendations”.

 

Regards,

 

Kyle Weeks, Ph.D.
Director of Product Strategy, SPSS Statistics Product Line

SPSS Inc.

Phone: [phone] / Fax: [phone]

<a href="BLOCKED::mailto:[e-mail]" title="mailto:[e-mail]">kweeks@...

www.spss.com
SPSS Inc. helps organizations turn data into insight through predictive analytics.

 

 

---

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Art Kendall
Sent: Monday, March 09, 2009 10:22 AM
To: [hidden email]
Subject: Re: SAS Vs SPSS

 

A lot depends on the particular situation.
SQL etc is external to either package. speed of access, number of users, remote connection. is very much more  a function of your network and platform not the stat software.

For large files where only one pass is required SPSS used to have a big advantage because it only created the scratch files when thy were going to be reread.  I do not know if SAS has caught up on this or not.

A major consideration is the total cost of ownership, purchase, user salary, machine resources, etc.

Purchase and license costs are often small relative to hardware, networking etc. Combined the purchase and platform costs are often small relative to people costs.

In contexts where people and calendar time are a large part of the consideration, SPSS has long had an advantage.  SPSS syntax is  more user friendly.  It is very consistent in its usage.  This tend to result in less time to write the syntax, "reference it" (audit-speak for the process of another person going over the syntax to see if it appears to do what it is supposed to), refine/maintain it, communicate to help sources, etc.  The syntax advantages are mostly in the part of the process that takes up most of the time on a project, assembling, cleaning, and other preparation of the data for analysis.

Of course, the portion of total cost that is for people, will be less of a consideration in some environments where a process is set up and then run as is for a long time doing a few things.  People costs would tend to be larger when there are many different ad hoc things that need to be done by a wide variety of people and smaller when a few people people do pretty much the same thing.


If you have the opportunity get syntax to do data cleaning, preparation, etc. tasks, i.e.,  a lot of data transformations, recodes, autorecodes, scoring, right/wrong marking, etc from both both packages.
Show both to some people who use neither package.  See how well they can tell what the syntax is doing.

By way of disclosure my experience is mostly on the kind of applications where researchers/scientist/faculty/analysts  use stat software intermittently.  Much less of it is on the kind of things like payroll, inventory, etc.  where a few people do a small range of things.

HTH
Art Kendall
Social Research Consultants

Krishna Chaitanya wrote:

Hi SPSSers,

 

I am Krishna working as a consultant. I am comparing SAS server package with the SPSS Server for one of my clients. I have some basic doubts on how SPSS fares with SAS on the below topics --

Ability to extract data from an Oracle/ACCESS database

Handling huge amounts of data (Is there an alternate to proc sql and proc transpose)

Processing time

Opening multiple sessions (Will 15 people be able to connect simultaneously?)

Remote connection (How fast is this?)

Cost – Long term and short term

 

Any help in this topic is highly appreciated.

 

Thanks,
Krishna

 

***********************************************************************

This e-mail and any files transmitted with it are for the sole use of the intended recipient(s) and may contain confidential and privileged information. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies and the original message. Any unauthorized review, use, disclosure, dissemination, forwarding, printing or copying of this email or any action taken in reliance on this e-mail is strictly prohibited and may be unlawful.

 The recipient acknowledges that Mu Sigma or its subsidiaries and associated companies (collectively " Mu Sigma Inc), are unable to exercise control or ensure or guarantee the integrity of over the contents of the information contained in e-mail transmissions and further acknowledges that any views expressed in this message are those of the individual sender and no binding nature of the message shall be implied or assumed unless the sender does so expressly with due authority of Mu Sigma. Before opening any attachments please check them for viruses and defects.

 ***********************************************************************