spssinc modify tables - deleting rows

[jagadishpchary]

Hi,

I am using the below syntax in spss to delete some rows in the tables (column proportions table as well).
unfortunately i tried to modify the code but noluck.

spssinc modify tables subtype="custom table" Select ="L2B" dimension=rows
level=-2 process=preceding hide=true.


Could anyone please help me in correcting the code.

Attached the required table format in the attached excel report i.e.
table-2.
Book4.xlsx
<http://spssx-discussion.1045642.n5.nabble.com/file/t340698/Book4.xlsx> 

Also attached the spss data file.
data.sav
<http://spssx-discussion.1045642.n5.nabble.com/file/t340698/data.sav> 

thanks for the help.

Regards,
Jagadish



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
To manage your subscription to SPSSX-L, send a message to
[e-mail] (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]

The level parameter should be -1 since you are hiding the rows with the innermost specified label.

On Sat, Nov 23, 2019 at 8:49 AM jagadishpchary <[hidden email]> wrote:

Hi,

I am using the below syntax in spss to delete some rows in the table.
unfortunately i tried to modify the code but noluck.

spssinc modify tables subtype="custom table" Select ="L2B" dimension=rows
level=-2 process=preceding hide=true.


Could anyone please help me in correcting the code.

Attached the required table format in the attached excel report i.e.
table-2.
Book4.xlsx
<http://spssx-discussion.1045642.n5.nabble.com/file/t340698/Book4.xlsx> 

Also attached the spss data file.
data.sav
<http://spssx-discussion.1045642.n5.nabble.com/file/t340698/data.sav> 

thanks for the help.

Regards,
Jagadish



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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

[jagadishpchary]

Hi Jon,

thanks for the reply.

I was unable to delete the mentioned raw in the main table by using -1. it
is getting deleted in the column proportions table only.

Regards,
Jagadish



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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]

That means that you are specifying the wrong table type to SPSSINC MODIFY TABLES.  The main table has a type of "customtable" while the column proportions type is 'Comparisons of Proportions'

On Mon, Nov 25, 2019 at 12:25 AM jagadishpchary <[hidden email]> wrote:

Hi Jon,

thanks for the reply.

I was unable to delete the mentioned raw in the main table by using -1. it
is getting deleted in the column proportions table only.

Regards,
Jagadish



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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

[jagadishpchary]

Hi Jon,

As said, i have used the below syntax. however the issue is not resolved.
May i know where i am going wrong?

spssinc modify tables subtype="customtable" Select ="L2B" dimension=rows
level=-1 process=preceding hide=true.
spssinc modify tables subtype="Comparisons of Proportions" Select ="L2B"
dimension=rows level=-1 process=preceding hide=true.

Regards,
Jagadish



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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

[Hermann]

Hi Jagadish,

dealing with the same problem since a lot of days.

The first MODIFY TABLES works perfect, the second does not. From my point of
view the problem is in adressing the area of a table which is different. I
was not able to find a solution for hiding instead of coloring the rows
addressed.

SPSSINC MODIFY TABLES
  SUBTYPE="Correlations"
  SELECT="Sig. (2-tailed)" "N"
  DIMENSION=ROWS
  PROCESS=PRECEDING
  /STYLES BACKGROUNDCOLOR=0 191 255.

SPSSINC MODIFY TABLES
  SUBTYPE="Correlations"
  SELECT="Sig. (2-tailed)" "N"
  DIMENSION=ROWS
  PROCESS=PRECEDING
  HIDE=TRUE.

Best regards, Hermann



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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]

Something has changed in the table structure or api behavior, but you can still achieve most of the same effect with this code.

SPSSINC MODIFY TABLES SUBTYPE='"Correlations"' SELECT=0 DIMENSION=ROWS PROCESS=PRECEDING
/styles customfunction=
"formatcorrmat.cleancorr(hidecounts=False, hidel=True, lowertri=False, hideinsig=1., emphasis=1., "+
    ")".

This appears in the menus as Utilities > Format Correlation Matrix

On Tue, Jan 28, 2020 at 1:33 AM Hermann <[hidden email]> wrote:

Hi Jagadish,

dealing with the same problem since a lot of days.

The first MODIFY TABLES works perfect, the second does not. From my point of
view the problem is in adressing the area of a table which is different. I
was not able to find a solution for hiding instead of coloring the rows
addressed.

SPSSINC MODIFY TABLES
  SUBTYPE="Correlations"
  SELECT="Sig. (2-tailed)" "N"
  DIMENSION=ROWS
  PROCESS=PRECEDING
  /STYLES BACKGROUNDCOLOR=0 191 255.

SPSSINC MODIFY TABLES
  SUBTYPE="Correlations"
  SELECT="Sig. (2-tailed)" "N"
  DIMENSION=ROWS
  PROCESS=PRECEDING
  HIDE=TRUE.

Best regards, Hermann



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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

[Hermann]

Thanks Jon,

but the syntax provided did not hide the significance statistic. Pearson
correlation only should remain in the table.

ATB, Hermann



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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]

This worked for me (generated from the dialog box).

CORRELATIONS
  /VARIABLES=jobtime minority prevexp salary
  /PRINT=TWOTAIL NOSIG
  /MISSING=PAIRWISE.

SPSSINC MODIFY TABLES SUBTYPE='"Correlations"' SELECT=0 DIMENSION=ROWS PROCESS=PRECEDING
/styles customfunction=
"formatcorrmat.cleancorr(hidecounts=True, hidel=True, lowertri=True, hideinsig=.05, emphasis=1., )".    

On Wed, Jan 29, 2020 at 1:03 AM Hermann <[hidden email]> wrote:

Thanks Jon,

but the syntax provided did not hide the significance statistic. Pearson
correlation only should remain in the table.

ATB, Hermann



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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

[Hermann]

The problem was that hiding insignificant correlations and significance
levels are joined together.

If I want to show all correlations including 0 that is on a p-level of
1.0000, then all rows for "Sig. (2-tailed) are displayed. If reducing the
parameter from 1 to "hideinsig=.99999" then the 0-correlations significant
on a p-level of 1.0000 are removed together with "Sig. (2-tailed)".

But I can work with.

Thanks



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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]

If I understand what you want, this syntax should do it.

SPSSINC MODIFY TABLES SUBTYPE='"Correlations"' SELECT=0 DIMENSION=ROWS PROCESS=PRECEDING
/styles customfunction=
"formatcorrmat.cleancorr(hiden=False, hidel=True, lowertri=False, hideinsig=1., emphasis=1., "+
    ")".

The dialog box puts together hiding insignificant correlations and the sig level rows, but they are handled separately  in the syntax.  I noticed, though, that the count hiding choice in the dialog generates the wrong keyword.  Instead of hidecounts, it should be hiden.  I edited the syntax above to be correct.  But I have also modified the formatcorrmat.py file to accept either keyword (but not both).  The updated .py file is attached.

If you want to update your system with this, find where formatcorrmat.py is saved and replace it with the attached file.  The possible locations can be displayed  by running SHOW EXT from a syntax window and looking in the EXTPATHS EXTENSIONS section of the output.  Restart Statistics after replacing the file.

If you are still not getting what you wanted, send me the syntax you are using.

On Thu, Jan 30, 2020 at 3:21 AM Hermann <[hidden email]> wrote:

The problem was that hiding insignificant correlations and significance
levels are joined together.

If I want to show all correlations including 0 that is on a p-level of
1.0000, then all rows for "Sig. (2-tailed) are displayed. If reducing the
parameter from 1 to "hideinsig=.99999" then the 0-correlations significant
on a p-level of 1.0000 are removed together with "Sig. (2-tailed)".

But I can work with.

Thanks



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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

[Hermann]

Thanks Jon,

is it possible to submit the python file with the txt-extension. Outlook has
blocked access to "formatcorrmat.py" due to administrator settings. I will
then rename it.

ATB, Hermann



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

=====================
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