OUTPUT MODIFY oddities

[Albert-Jan Roskam-3]

Hi,

In example 1 below, I am using an example I found under OUTPUT MODIFY in the Syntax Reference Manual. It is supposed to delete all output, except pivot tables. Well, it doesn't. It deletes everything. Am I doing something wrong?

Also, notice the inconsistent notation compared to OMS (example 2 works as expected btw): parentheses vs square brackets, without vs with equals sign, delete=yes vs viewer=no. 

* example 1.

get file = "samples\English\Employee data.sav".
frequencies gender.
echo "I will be deleted!".
output modify /select all except (tables) /deleteobject delete = yes.

* example 2.

get file = "samples\English\Employee data.sav".
oms /select all except = [tables] /destination viewer=no.
frequencies gender.
echo "I will be deleted!".
omsend.

Thanks!

Albert-Jan

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[Jon Peck]

Trying example 1 in V23 (with the latest fixpack), it works correctly.  What version are you using?

As for the oms viewer= vs the output modify delete=, for  oms, the default is yes, since that is what is usually wanted, in addition to which, a default of no would probably be confusing.  For output modify, since the delete keyword could well be combined with other formatting actions in the same command, having delete=yes as a default would wipe out the objects unless the user took the extra step of adding a deleteobject subcommand with delete=no just to prevent that.  It would be awkward for the delete action default to depend on whether deleteobject was issued.

For the other inconsistencies, they stem from the introduction of the "universal parser" some time after oms was created.  oms was introduced in SPSS 12.  I would have preferred more consistency in output modify with oms, but the UP did not support that, at least at that time.

On Wed, Aug 31, 2016 at 9:10 AM, Albert-Jan Roskam <[hidden email]> wrote:

Hi,

In example 1 below, I am using an example I found under OUTPUT MODIFY in the Syntax Reference Manual. It is supposed to delete all output, except pivot tables. Well, it doesn't. It deletes everything. Am I doing something wrong?

Also, notice the inconsistent notation compared to OMS (example 2 works as expected btw): parentheses vs square brackets, without vs with equals sign, delete=yes vs viewer=no. 

* example 1.

get file = "samples\English\Employee data.sav".
frequencies gender.
echo "I will be deleted!".
output modify /select all except (tables) /deleteobject delete = yes.

* example 2.

get file = "samples\English\Employee data.sav".
oms /select all except = [tables] /destination viewer=no.
frequencies gender.
echo "I will be deleted!".
omsend.

Thanks!

Albert-Jan

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

--

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[Albert-Jan Roskam-3]

Hi Jon,

I am on Spss v22 Fixpack 2. Unlike it's predecessor SpssInc Modify Output, Output Modify is not implemented using Python, right? Today I noticed that Python programmability (but not scripting) is broken in my installation.

Instead of viewer= or delete=, drop= would also have been nice. But I agree that the defaults should differ between the two commands.

I like Python's DeprecationWarnings. They allow the developer to improve things without immediately breaking any code. SPSS could have a SET command that controls whether similar warnings are displayed or not.

Best wishes,
Albert-Jan

---

From: SPSSX(r) Discussion <[hidden email]> on behalf of Jon Peck <[hidden email]>
Sent: Wednesday, August 31, 2016 3:50:40 PM
To: [hidden email]
Subject: Re: [SPSSX-L] OUTPUT MODIFY oddities

 

Trying example 1 in V23 (with the latest fixpack), it works correctly.  What version are you using?

As for the oms viewer= vs the output modify delete=, for  oms, the default is yes, since that is what is usually wanted, in addition to which, a default of no would probably be confusing.  For output modify, since the delete keyword could well be combined with other formatting actions in the same command, having delete=yes as a default would wipe out the objects unless the user took the extra step of adding a deleteobject subcommand with delete=no just to prevent that.  It would be awkward for the delete action default to depend on whether deleteobject was issued.

For the other inconsistencies, they stem from the introduction of the "universal parser" some time after oms was created.  oms was introduced in SPSS 12.  I would have preferred more consistency in output modify with oms, but the UP did not support that, at least at that time.

On Wed, Aug 31, 2016 at 9:10 AM, Albert-Jan Roskam <[hidden email]> wrote:

Hi,

In example 1 below, I am using an example I found under OUTPUT MODIFY in the Syntax Reference Manual. It is supposed to delete all output, except pivot tables. Well, it doesn't. It deletes everything. Am I doing something wrong?

Also, notice the inconsistent notation compared to OMS (example 2 works as expected btw): parentheses vs square brackets, without vs with equals sign, delete=yes vs viewer=no. 

* example 1.

get file = "samples\English\Employee data.sav".
frequencies gender.
echo "I will be deleted!".
output modify /select all except (tables) /deleteobject delete = yes.

* example 2.

get file = "samples\English\Employee data.sav".
oms /select all except = [tables] /destination viewer=no.
frequencies gender.
echo "I will be deleted!".
omsend.

Thanks!

Albert-Jan

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

--

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

This is a bug in v22 I reported in this blog some time ago, if I remember correctly.

It was corrected in v23.

Greetings,

  Mario

Albert-Jan Roskam <[hidden email]> schrieb am 19:21 Mittwoch, 31.August 2016:

Hi Jon,

I am on Spss v22 Fixpack 2. Unlike it's predecessor SpssInc Modify Output, Output Modify is not implemented using Python, right? Today I noticed that Python programmability (but not scripting) is broken in my installation.

Instead of viewer= or delete=, drop= would also have been nice. But I agree that the defaults should differ between the two commands.

I like Python's DeprecationWarnings. They allow the developer to improve things without immediately breaking any code. SPSS could have a SET command that controls whether similar warnings are displayed or not.

Best wishes,
Albert-Jan

---

From: SPSSX(r) Discussion <[hidden email]> on behalf of Jon Peck <[hidden email]>
Sent: Wednesday, August 31, 2016 3:50:40 PM
To: [hidden email]
Subject: Re: [SPSSX-L] OUTPUT MODIFY oddities

 

Trying example 1 in V23 (with the latest fixpack), it works correctly.  What version are you using?

As for the oms viewer= vs the output modify delete=, for  oms, the default is yes, since that is what is usually wanted, in addition to which, a default of no would probably be confusing.  For output modify, since the delete keyword could well be combined with other formatting actions in the same command, having delete=yes as a default would wipe out the objects unless the user took the extra step of adding a deleteobject subcommand with delete=no just to prevent that.  It would be awkward for the delete action default to depend on whether deleteobject was issued.

For the other inconsistencies, they stem from the introduction of the "universal parser" some time after oms was created.  oms was introduced in SPSS 12.  I would have preferred more consistency in output modify with oms, but the UP did not support that, at least at that time.

On Wed, Aug 31, 2016 at 9:10 AM, Albert-Jan Roskam <[hidden email]> wrote:

Hi,

In example 1 below, I am using an example I found under OUTPUT MODIFY in the Syntax Reference Manual. It is supposed to delete all output, except pivot tables. Well, it doesn't. It deletes everything. Am I doing something wrong?

Also, notice the inconsistent notation compared to OMS (example 2 works as expected btw): parentheses vs square brackets, without vs with equals sign, delete=yes vs viewer=no. 

* example 1.

get file = "samples\English\Employee data.sav".
frequencies gender.
echo "I will be deleted!".
output modify /select all except (tables) /deleteobject delete = yes.

* example 2.

get file = "samples\English\Employee data.sav".
oms /select all except = [tables] /destination viewer=no.
frequencies gender.
echo "I will be deleted!".
omsend.

Thanks!

Albert-Jan

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

--

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

Albert-Jan, if I remeber correctly you can fix it by using

/select all except (tables outlineheaders)

in v22.

Greetings,

  Mario

Mario Giesel <[hidden email]> schrieb am 12:53 Donnerstag, 1.September 2016:

This is a bug in v22 I reported in this blog some time ago, if I remember correctly.

It was corrected in v23.

Greetings,

  Mario

Albert-Jan Roskam <[hidden email]> schrieb am 19:21 Mittwoch, 31.August 2016:

Hi Jon,

I am on Spss v22 Fixpack 2. Unlike it's predecessor SpssInc Modify Output, Output Modify is not implemented using Python, right? Today I noticed that Python programmability (but not scripting) is broken in my installation.

Instead of viewer= or delete=, drop= would also have been nice. But I agree that the defaults should differ between the two commands.

I like Python's DeprecationWarnings. They allow the developer to improve things without immediately breaking any code. SPSS could have a SET command that controls whether similar warnings are displayed or not.

Best wishes,
Albert-Jan

---

From: SPSSX(r) Discussion <[hidden email]> on behalf of Jon Peck <[hidden email]>
Sent: Wednesday, August 31, 2016 3:50:40 PM
To: [hidden email]
Subject: Re: [SPSSX-L] OUTPUT MODIFY oddities

 

Trying example 1 in V23 (with the latest fixpack), it works correctly.  What version are you using?

As for the oms viewer= vs the output modify delete=, for  oms, the default is yes, since that is what is usually wanted, in addition to which, a default of no would probably be confusing.  For output modify, since the delete keyword could well be combined with other formatting actions in the same command, having delete=yes as a default would wipe out the objects unless the user took the extra step of adding a deleteobject subcommand with delete=no just to prevent that.  It would be awkward for the delete action default to depend on whether deleteobject was issued.

For the other inconsistencies, they stem from the introduction of the "universal parser" some time after oms was created.  oms was introduced in SPSS 12.  I would have preferred more consistency in output modify with oms, but the UP did not support that, at least at that time.

On Wed, Aug 31, 2016 at 9:10 AM, Albert-Jan Roskam <[hidden email]> wrote:

Hi,

In example 1 below, I am using an example I found under OUTPUT MODIFY in the Syntax Reference Manual. It is supposed to delete all output, except pivot tables. Well, it doesn't. It deletes everything. Am I doing something wrong?

Also, notice the inconsistent notation compared to OMS (example 2 works as expected btw): parentheses vs square brackets, without vs with equals sign, delete=yes vs viewer=no. 

* example 1.

get file = "samples\English\Employee data.sav".
frequencies gender.
echo "I will be deleted!".
output modify /select all except (tables) /deleteobject delete = yes.

* example 2.

get file = "samples\English\Employee data.sav".
oms /select all except = [tables] /destination viewer=no.
frequencies gender.
echo "I will be deleted!".
omsend.

Thanks!

Albert-Jan

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

--

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD