Learning to use lookup transformation - match value in one of two columns

Im a little new to using lookup transforms. Been reading a bit, but using two columns is throwing me off.
Have a dataflow with the source having a column I want to use to lookup in a table in another DB. The match could be in one of two columns in that table.
If a match is found, I need a different value from the lookup table added to the source flow that is going into the destination.
How can I match on one of two columns in the lookup table, and use a totally different column added to the final input going into the destination ?
In the Lookup configuration, I can set the lookup operation to replacing the value in the source, but I want EITHER value match from the two columns from the lookup table to go into the value that will go into the destination.
With TSQL - I would use a case statement and compare the source column to BOTH of the lookup table columns.

You can handle this within single lookup using below query
SELECT CASE WHEN <condn1> THEN Field1 ELSE Field2 END AS lookupField,
CASE WHEN <condn1> THEN Field3 ELSE Field4 END AS MatchField
FROM Table2
Then just do single relationship with LookupField and select MatchField to be aded to output
the condition will specify your fields value condition based on which you determine which column to be matched against
Please Mark This As Answer if it solved your issue
Please Mark This As Helpful if it helps to solve your issue
Visakh
My MSDN Page
My Personal Blog
My Facebook Page

Similar Messages

Using Excel Functions, such as VLOOKUP to compare two columns of text for similarities

My problem seems simple but I am struggling to figure it out. I have two columns of text. I am trying to see if the text in column B appears in column A. However, it is not always an exact match, and I basically want to check if column A contains the text
that appears in column B. What is the best way to do this?

In C2, use a formula like
=IF(SUMPRODUCT(NOT(ISERROR(SEARCH(B2,$A$1:$A$1000)))*1)>0,"Found","Not Found")
then copy down to match column B.
This formula will return "Found" when the entire contents a cell in B is found somewhere in any cell in A1:A1000. E.G., "red" in B2 is found in cell A100 which contains "I have a red dog"

Is it possible to do lookup without using lookup transformation?

Could you please brief about data masking transformation..?

Hi Deepsikha, You can also use Data Masking Transformation (Substitution Masking) in place of lookup. This performs similar functionality as Lookup. Thanks,Ramesh.

Not Working-LexicalParamater used to send multiple values with one prameter

Hi,
I am calling a reportname.rdf from the web(using apache webserver and url as follows:
http://localhost/cgi-bin/rwcgi60.exe?server=reportservername&report=acctActivity.rdf&userid=bizsystest/[email protected]&destype=cache&desformat=pdf&placeholdercolumnname='2,3,4'&docid=2
I have to send multiple values for fundid.
To accomplish this,
I created a placeholder column and initialized the value '0' and type taken as character(even though datatype of this column is number).
In sql query
select table1.c1, table2.c2 from table1, table2, table3, table4 where table1.cl in (&placeholdercolumname)
and table1.c2 = table2.c2(+)
and table3.doccode = :docid
From Web, URL contains the placeholdercolumnnmae='2,3,4'
Still its taking the value zero and output is generating.
Its not taking my newvalues ('2,3,4') in IN parameter.
I donot know the reason, where exactly is wrong.
Please help me to resolve.
2. My Second Issue:
Is there anyway, can we hide all key/value parameters send through the URL not visible to endusers
like servername, database userid and password and configure some where, so system can pickup each time when I run report.
Please help me to resolve.
Thanks,
Usha.

Hi Usha,
First of all, I assume that the place holder column you have specified is a UserParameter which you have specified on the command line. I don't think there is any direct way of doing that.
I created a placeholder column and initialized the value '0' and type taken as character(even though datatype of this column is number). If I infer correctly from what you are saying, you can't mix and match the datatypes of the columns, you can't define the parameter as a number and pass characters. I suggest you declare your parameter as a character with sufficient width to hold your LOV and then process them convert them to numbers.
Another suggestion is try replacing your "Placeholder" with another query construct.
For your Second issue,
You can use the Key-Map file Oracle Reports provides for the Web client. Please refer to
<ORACLE_HOME>\Report60\server\cgicmd.dat.
In the cgicmd.dat,you can alias your entire commandline with a single key and use that key in your URL.
Regards
Oracle Reports Team.

Using case pull infor from one of two columns

how can I use the case statement to pull data from rows that are grouped together from a table? or is there a way to use if..then?
my incorrect , unfinished query.....
select t1.col1 from tab1 t1,tab2 t2
where
t1.col1= t2.col1 and
t1.col1= < if I find an 'x' in t2.col2 then return all t1 rows from that group that have x's in corresponding t2.col2
else return all rows that have an x's in corresponding t2.col3>
group by t1.col1

One table has data like this:
STG_CLAIMANT_XREF
SRC_EVENT_NUMBER     SRC_CLAIM_NUMBER     SRC_CLMT_NO     STG_CLAIM_TYPE
0040283362              00402833620001                 0001          112
0040283362              00402833620002                 0002          111
0040283362              00402833620003                 0003          112
0040283362              00402833620050                 0050          115The above represents one group of data from a group by SRC_EVENT_NUMBER,
a subset of groups from the whole table.
The other has data like this:
CLAIMTYPE_X_BI_PD
CODE_ID     PD_TYPE     BI_TYPE
110       -1
111            -1
112       -1
119            -1I want to run through the 1st table and match the STG_CLAIM_TYPEs to the CODE_IDs
in the second table.
I'm going to join STG_CLAIM_TYPE = CODE_ID.
If ANY rows in the first table have STG_CLAIM_TYPEs with values in the PD_TYPE col in the second table than for that group I can ONLY return the rows that have STG_CLAIM_TYPEs = 111 for example.
However, if No rows of type 111 exist but rows STG_CLAIM_TYPEs = 111 exist (STG_CLAIM_TYPEs with values in the BI_TYPE col) Than I can only return rows that have STG_CLAIM_TYPEs = 112
my first try was a query that worked but returned both type out of each group. It was
select xr.SRC_EVENT_NUMBER, xr.SRC_CLAIM_NUMBER, xr.STG_CLAIM_TYPE
from STG_CLAIMANT_XREF xr, CLAIMTYPE_X_BI_PD BP
where
xr.STG_CLAIM_TYPE = bp.CODE_ID
and
(bp.PD_TYPE = -1 or bp.BI_TYPE = -1)
group by xr.SRC_EVENT_NUMBER, xr.SRC_CLAIM_NUMBER, xr.STG_CLAIM_TYPE;
will return rows 1,2,& 3. I need only 1 & 3 or just 2.
I figured using a case statement would help me solve this issue but I am usure of how to
implement it.
What would you suggest?
Message was edited by:
user623359

How to get the value from one Popup lov column to another popup lov column

Hi,
I am new to oracle apex development and having the below issue.
In my application, there is a tabular form with 15 columns ( c1.. c15).
I have to populate the value of column C5 based on the selected(from popup lov) value of column C3, tried to use onchange, but didn't help much.
Any help please.
Thanks and Regards,

Oh boy, this is a fun one.
onchange should work theoretically (in this example, assume that f05 is the target column that should be set and "this" is the source item whose value is to be transferred to f05 on the same row (row 2)):
onchange=$s('f05_0002',$v(this));
BUT the catch is of course that needs to be different for every row (can't hardcode the '2'), so you need something to dynamically create the row number component.
I wrote this for an app I'm working on that uses master-detail forms heavily (I also wrote a lot more code to read the fmap array that is in v4 so that I can reference my cells via their column name and not the numeric position (so "f05 can be determined w/o hard coding), insulating against columns moving around, columns being made display-only etc. but I won't bore you with that here unless you really need to know).
function getRow(pObj)
{ //Pass in an object reference to a tabular form cell and get back an integer
//derived from the object ID.
var vRow=pObj.id.substr(pObj.id.indexOf("_")+1);
if (isNaN(vRow))
    return (null);
return (parseInt(vRow,10));
function formatRow(pRow)
{ //Pass in an integer and it'll be returned as the tabular form cell ID suffix
//(e.g.: pass in 1 and get back string "_0001").
//Used in building ID references to tabular form cells.
if((isThingEmpty(pRow)) || (isNaN(pRow)))
    return(null);
var vRow=pRow.toString();
while(vRow.length<4)
    vRow="0"+vRow;
return("_"+vRow);
}Therefore:
onchange=$s('f05_'+formatRow(getRow(this)),$v(this));
So in essence, pass in "this" which will be a reference to the current item, largely to determine what row we're on. getRow will return 1, 2, 3, etc. formatRow will take that and return 0001, 0002, 0003, etc. Once that is done, it'll concatenate with the f05 and f04 to refer to the correct columns (determining f05, f04, etc. dynamically is another matter if you really need to but I didn't want to complicate this answer too much at once).
Note: above I also use a isThingEmpty() function that I wrote. It does nothing other than check the item for an empty string, if the item is null, etc. Just do your own evaluation of empty-string, no-value, etc. there.
It would indeed be nice though if Apex had a better way to delclaratively access the tabular form items though. Despite all the v4 enhancements, tabular forms were not entirely upgraded to the same full functionality of page items.

Oracle Procedure- Get aggrigate value based on comparing two columns

Hi All,
I have written a Procedure which accepts one input parameter and returns one refcursor.
The logic used in the procedure is:
The procedure is used to display summary as well as detail report. When the input parameter have more than one comma saperated values then it will act as a summary report and when there will be just one value as the input then it will act as a detail report.
The input parameter is like 'AA12345,BB45434,HJ89736' the first two character is Brand_id and the rest of the 5 character is the propery_id.
The aggrigation aggrigation query is :
Select Month_name,Year,sum(REVPAR),Sum(ADR) from FIS_KPI_TREND where
Brand_id='AA' and SIte_id='12345' group by Month_name,Year;
The above query will worke fine for the Detail report when there will be just one value for the input as 'AA12345'.
The problem is comming for the Summary report when I need to get the aggrigated value for all the brand and site which came in the input parameter.
So i changed the query as:
Select Month_name,Year,sum(REVPAR),Sum(ADR) from FIS_KPI_TREND where
Brand_id||SIte_id in ('AA12345','BB45434','HJ89736') group by Month_name,Year;
(I'm using parser function to parse comma saperated values.)
By doing this i'm getting the desired result but the performance has decrised as the Index is based on Brand_id and Site_id but i'm Using Brand_ID||Site_id for my search.
Can any one suggest any other way to rewrite this query so that it can take the index as well as give me the sesired output.
Database used : Oracle 10g
Thanks in advance.
Sumit Narayan
Edited by: Sumit Narayan on Apr 11, 2010 10:00 PM

I'd be inclined to do something like this to parse the comma separated list, which would allow you to join without concatenating columns like you're doing now.
ME_XE?select
substr(column_value, 1, 2)            as brand_id,
substr(column_value, 3, length(column_value))   as site_id
from table(sys.ODCIVARCHAR2LIST('AA12345','BB45434','HJ89736'));
2    3    4
BRAND_ID   SITE_ID
AA         12345
BB         45434
HJ         89736
3 rows selected.
Elapsed: 00:00:00.20
ME_XE?
ME_XE?select * from v$version;
BANNER
Oracle Database 10g Express Edition Release 10.2.0.1.0 - Product
PL/SQL Release 10.2.0.1.0 - Production
CORE 10.2.0.1.0 Production
TNS for Linux: Version 10.2.0.1.0 - Production
NLSRTL Version 10.2.0.1.0 - Production
5 rows selected.
Elapsed: 00:00:00.39
ME_XE?So for your case ...
Select
f.Month_name,
f.Year,
sum(f.REVPAR),
Sum(f.ADR)
from FIS_KPI_TREND f, table(sys.ODCIVARCHAR2LIST('AA12345','BB45434','HJ89736')) t
where f.Brand_id = t.brand_id
and   f.SIte_id = t.site_id
group by f.Month_name, f.Year;Edited by: Tubby on Apr 11, 2010 10:12 PM

Merging values of CKFs in two columns into single column

Hi ,
I want to merge values present in two different columns.The values present in rows of one column are not present in rows of other column, I have to merge the values into a single column.so that in the single column i can get all the values.how to do it in BEx.
regards ,
man.

Hi,
Dear murali , the solution you gave has helped me a bit as i am able to get the values for both the formula KF's in the same column,But another problem has arisen.
   For both the formula KF's I had to choose 'New Selection' (and not 'New formula')after R-cllick on the structure, as I had to use a Characteristic as restriction for Each of the KF's, and also to use the Characteristic I had to associate it with some KF so that the Characteristic wil be used as a KF in the formula.
   Now what is happening is that I am getting the values of that associated KF also in the empty rows.(i.e. the rows where the value of both the KF1 and KF2 is not present)
   What has to be done to remove them and keep the empty rows empty and not show the value of the KF taken with the Characteristic for restriction. Also its value is summing up with KF1 and KF2 in the coulmn which is having both KF1 and KF2 together.
Regards
man

Prevent duplicate value being entered into a column

How do we prevent user to enter duplicate value into one of the column of a multiple record block before we commit the form ?
Can anybody help ?
Thanks,
Shu Wen

Do you want the duplicate to be detected as soon as the item is validated ?
If so, the only way I have found to do so is to maintain a copy of the contents of the block in a record group. Rather cumbersome and a few triggers needed, but I have one application where I had to do it. I have always found it restrictive that you cannot read values from other records without navigating to them. If anyone has found a better method I would be interested to hear.

How to suppress Lookup transformation no-match output if all rows match.

Hi. I have a Lookup Transformation that sends its matched output to an OLEDB destination and its no-match output to a flat file destination. The Lookup Transformation is configured to "Redirect rows to no match output" when there are no matches.
The problem is that even when all rows match (there is no no-match output), the flat file destination creates an empty file. How can I change this so that the no-match out put "runs" only if there is no-match output? I don't want an empty file to
be created if all rows match.
Thanks.

Hi Peter,
If I am not mistaken, there is no way of controlling the creation of empty files. However this can be handled by adding a "File System Task" to Control Flow for deleting the file if no records are gone through "No-Match Output". Here are the steps for doing
it;
Create a variable for holding number of records going through No-Match output.
Use the variable with "Row Count" transformation. Set "Row Count" in between "Look up" and "Flat File Destination".
Add a "File System Task" just after the "Data Flow Task" for deleting the file.
Set Precedence constraint with an expression using the variable created for executing "File System Task".
Dinesh Priyankara
http://dinesql.blogspot.com/
Please use Mark as answer (Or Propose as answer) or Vote as helpful if the post is useful.

Can VLOOKUP, OFFSET, INDEX and MATCH be used to SUM the value of cells from multiple tables?

I have a problem thats needs a sulution. I have a Spreadsheet with multiple sheets and tables. The (Truck ) Sheets each represent a "Site" and the tables within from 1-31 show inventory counts for each day of the month.
Shown below are Edited example's of a "Site" sheet and the Day tables contained within it.
The Spreadsheet expands as a new "Site" (Truck ) is added regularly. Currently there are 30 "Sites" (Trucks).
Also within the spreadsheet is a sheet (Checkpoint) which contains tables used to show summary data for each "Site" sheet. There is one table used to review one site, so currently there are 30 tables identical in format and structure reviewing 30 sites.
An edited example of the summary table for Truck 1 is shown below.
The table's in CP reference data within cells from other sheets. A LEN INDIRECT formula
=IF(LEN(INDIRECT("Truck 1::Day "&COLUMN(cell)−3&"::K53",addr-style))<1,"",INDIRECT("Truck 1::Day "&COLUMN(cell)−3&"::k53",addr-style))
was used to retrieve the values within certain cells. This formula works perfectly because when I add another "Site" I simply need to Dupicate a "Site" Template Sheet and then Copy Paste the LEN INDIRECT formula in to the newly created Summary table.
There is another sheet (Command Central) which contains a Master summary table. The Master Summary sheet is identical in structure to the CP Summary Tables. Its purpose is to show the combined data from the tables within Checkpoint (Truck 1-30). for instance MS::D5=Checkpoint::Truck 1:Truck 30::D5. I realize that Numbers cannot calculate cells within a range of table so I am hoping for a workaround.
Could someone please help me with a formula something like the LEN INDIRECT formula mentioned above or possible a something else that could work beside
D5=Truck 1::D5,Truck 2::D5,Truck 3::D5...?
When I use that formula type I will have to edit the formula every time I add a new site.
Anothe issue Im having is getting D3:AH3 to show the correct count of names. I am using the following formula:
D3=SUM(COUNTA(Truck 1::D3,Truck 2::D3,Truck 3::D3,Truck 4::D3))−SUM(COUNT(UNION.RANGES(Truck 1::D3,Truck 2::D3,Truck 3::D3,Truck 4::D3)))
The cells it is counting are either blank (the refenced cell contains a LEN INDIRECT formula that places "" if needed) or contain a name. This formula works when I and addressing cells without the LEN INDIRECT formula. What am I doing wrong?

Hello Wayne,
Here are screenshots of the tables in question. The first on is the Master summary table.
Cell D5 shows the value pulled from Checkpoint::Truck 1::D5
Cell E5 shows the value pulled from Checkpoint::Truck 2::D5
Cell F5 shows the value pulled from Checkpoint::Truck 3::D5
Cell G5 shows the value pulled from Checkpoint::Truck 4::D5

Lookup transformation to avoid duplicate rows? - SSIS 2005

Hi,
I'm maintaning a SSIS 2005 pkg. I need to read a flat file to write on a SQL Server table, avoiding duplicates.
I can have duplicates rows into the flat file to import and I need to prevent the insert of any rows already existing in the SQL Server table.
So, I think to use a lookup transformation. I've created a flat file source, then I connect it to a lookup transformation and inside it I've specified as the reference table the SQL Server destination table. Then, I've checked the available lookup columns
each adding as a new column: but the lookup task has arised an error and so I've specified as lookup operation the replacement. For each unmatching I need to write on the SQL Server table (the reference table in the lookup). For the lookup output error I've
indicate to ignore failure. Other steps?
However, when I run the pkg then inside the SQL Server destination table I can see only NULL values, but I want to see the rows don't already present in the table.
Any suggests to me, please? Thanks

Hi,
I'm using SSIS 2005 as reported in the title of the post.
I could have duplicates inside the source file and the existing table could haven't any rows.
Thanks
If you dont have any rows in existing table, then they will go through Error output in lookup task. For duplicates, lookup task will find matches and will go through lookup match output
Please Mark This As Answer if it helps to solve the issue Visakh ---------------------------- http://visakhm.blogspot.com/ https://www.facebook.com/VmBlogs

Lookup transformation - Performance Issue -Please help!

Hi,
I have a Source table with 5 million rows. I am pulling all the rows from the source table, then doing the lookup with 5 different tables one by one.
I use Full cache as the lookup table size is very less only. I used 'Ignore Failure' option in lookup table as I need to do the left join. That means, even if there is NO match, those records also should be passed to the bottom.
Now, the problem is it is taking a lot of time. To load 1,00,000 records, it is taking 1 hour. The how about 5 Million rows? 50 hours? Could you please help me to find out what is the mistake i am doing here?

Is the performance still very poor if you only add for example a row count transformation after your source?
How many rows are there in the lookup tables? Are you only selecting the columns you need? Is the data type of the selected columns very large?
Please mark the post as answered if it answers your question | My SSIS Blog:
http://microsoft-ssis.blogspot.com |
Twitter

Creating an XML From a Deep Structure using XSL Transformation

Hi ABAPers,
I have a requirement to use XSL Transformations on an ABAP deep type structure.
Currently i have an API that fills in this deep structure and by using CALL TRANSFORMATION ID.... i will get the BIG XML having having 100s of nodes . But actualy form the deep structure i need only some NODES (say 50)... So i tried writing an XSLT
in the transaction STRANS.. but on using this TRANSFORMATION which i wrote i am getting an error messgae like INVALID XML...
Am i going in right track or is there a good solution...
My sample transformation is as below...
<xsl:transform version="1.0"
xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
<xsl:strip-space elements="*"/>
<xsl:template match="/">
<xsl:value-of select="DATA/NODE_ELEMENTS/UUID_KEY/UUID"/>
<xsl:value-of select="DATA/NODE_ELEMENTS/SEMANTICAL_NAME"/>
<xsl:value-of select="DATA/NODE_ELEMENTS/STRUCT_CAT"/>
<xsl:value-of select="DATA/NODE_ELEMENTS/USAGE_CAT"/>
<xsl:value-of select="DATA/NODE_ELEMENTS/RESTRICTED_IND"/>
<xsl:value-of select="VALUES/DATA/NODE_ID"/>.
</xsl:template>
</xsl:transform>
Please help me in solving this issue....
Thanks,
Linda.

Hi Linda,
    I am replying based on your sample code.
   Try the below following suggestions.
   here 'GRPHDR' is the node where I am selecting the data.
           IGRPHDR is the name of the reference.
First calling the transformation in you program.
TYPES: BEGIN OF tl_hdr,
           msgid(20)    TYPE c,
             END OF tl_hdr.
DATA : t_hdr           TYPE STANDARD TABLE OF tl_hdr.
GET REFERENCE OF t_hdr INTO l_result_xml-value.
    l_result_xml-name = 'IGRPHDR'.
    APPEND l_result_xml TO t_result_xml.
   TRY.
        CALL TRANSFORMATION yfi_xml_read
        SOURCE XML it_xml_data
        RESULT (t_result_xml).
      CATCH cx_root INTO l_rif_ex.
        l_var_text = l_rif_ex->get_text( ).
        l_bapiret-type = 'E'.
        l_bapiret-message = l_var_text.
        APPEND l_bapiret TO errormsgs.
        EXIT.
    ENDTRY.
in XSL transformation
   First write a block of statement to specify from which node you are taking the data.
   No matter it is a node or sub-node.
<xsl:transform xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version="1.0">
<xsl:output encoding="iso-8859-1" indent="yes" method="xml" version="1.0"/>
<xsl:strip-space elements="*"/>
<xsl:template match="/">
      <asx:abap xmlns:asx="http://www.sap.com/abapxml" version="1.0">
      <asx:values>
        <IGRPHDR> " reference name of internal table
          <xsl:apply-templates select="//GrpHdr"/>
        </IGRPHDR>
</asx:values>
    </asx:abap>
</xsl:template>
Next select the data from the nodes under the nodes specified in the transformation.
here msgid is the field i am selecting for value.
<xsl:template match="GrpHdr">
    <item>
      <MSGID> " field in the internal table t_hdr where data has to go
        <xsl:value-of select="MsgId"/>
      </MSGID>
    </item>
</xsl:template>
reply back if further clarification is needed.
Thanks and regards,
Kannan N

Can we replicate Informatica Lookup transformation in ODI Interface?

So was just wondering if we could replicate a lookup transformation in an ODI join of some sort where in only one value is returned for multiple matches of the same item....something like a max for a group or something while joining 2 tables...
So for example I have 2 tables...A columns (Attendee_id,xyx,byx)
B columns (partnerid,attendee_id,xyz,bxc)
I want to join table A and B on Attendee_id and get the partner_id---but I want only one record in case of multiple matches for a particular attendee_id
can we do this??

No responses yet...Ive tried doing a left outer join...but it still returns all the matches...I want to return only one record....ne help from the xperts?

Learning to use lookup transformation - match value in one of two columns

Similar Messages

Maybe you are looking for