Thank you for your help Arek.

So to get this to work, the datasets must be in different databases?

And by the looks of your XML I don't get any of these (pc05) etc pointers? Is there because it's not a partition now? Sorry for questions, just trying to understand.

Cheers &  Merry Xmas to you too :-)

Ellen.

--
[email protected]
JDRF/WT Diabetes and Inflammation Laboratory (DIL)
Cambridge Institute for Medical Research (CIMR)
Wellcome Trust/MRC Building
Addenbrooke's Hospital Hills Road Cambridge CB2 0XY
Tel: 01223 763215

On 21/12/12 11:38, Arek Kasprzyk wrote:
Hi Ellen,
I understand now.

Here is what you have to do:


1. Create separate database for each dataset (e.g hs_130, hs_137, etc)
2. Make sure the table names are identical in each database (eg ... marker__main).
3. Open MConfigurator 'add source' (eg 'hs_130)
4. Right panel, 'data source management'. 'add' (eg hs_137)
5. Update data source
6. Add access point
7. Start the server.

I checked it properly this time (inserting different values to different databases) and everything seemed to be querying ok

Below are examples that I used (please do not get confused with the use of 'marker_130' table names in both databases, I was just too lazy to edit it properly but principle remains the same: identical names)


+-------------------------------+
| Tables_in_ellen_130           |
+-------------------------------+
| hs_dbsnp130__markers__main    |
| hs_dbsnp130__markers_func__dm |
+-------------------------------+

+-------------------------------+
| Tables_in_ellen_137           |
+-------------------------------+
| hs_dbsnp130__markers__main    |
| hs_dbsnp130__markers_func__dm |
+-------------------------------+
2 rows in set (0.00 sec)



(I am sending an example XML working with these tables in a separate email)

Please let me know if this solves the problem

Merry Xmas :)

a






On Thu, Dec 20, 2012 at 11:10 AM, Ellen Schofield <[email protected] <mailto:[email protected]>> wrote:

    Hi Arek,

    Many thanks for your help, sorry I've not got back to you before
    now, other work took over :-)

    I have loaded up your xml, tweaked to point at my db to see how it
    works.  All starts ok etc, but when I select the dbSNP137 dataset,
    it is still searching against the dbSNP130 data.

An example from the BioMart Portal could be the ensembl one. There is a single starting point (ensembl) and then multiple
    datasets (humna, mouse etc) that you can choose, all, I guess
    witht he same structure behind the scenes and the same queries etc.

    Thanks,

    Ellen.

-- [email protected] <mailto:[email protected]>
    JDRF/WT Diabetes and Inflammation Laboratory (DIL)
    Cambridge Institute for Medical Research (CIMR)
    Wellcome Trust/MRC Building
    Addenbrooke's Hospital Hills Road Cambridge CB2 0XY
    Tel:01223 763215  <tel:01223%20763215>

    On 12/12/12 17:40, Arek Kasprzyk wrote:
    Hi Ellen,
    I am sending my XML file. Please let me know if this is what you
    need or your problem is different

    a


    On Wed, Dec 12, 2012 at 2:33 PM, Arek Kasprzyk
    <[email protected] <mailto:[email protected]>> wrote:

        Hi Ellen,
        perhaps I did not understand what you are after. If you
        partition a table it will create many more smaller tables. I
        assumed that you were looking for the opposite effect to
        somehow "group" your tables? How do you want your datasets to
        look through GUI? Could you give me an example either from
        the BioPortal or from CancerPortal?

        cheers,
        a


        On Wed, Dec 12, 2012 at 12:02 PM, Ellen Schofield
        <[email protected]
        <mailto:[email protected]>> wrote:

            Hi Arek,

            Many thanks for your time.  Good to know that the
            structure is OK.

            These are the steps that I did...
            Imported as a relation mart
            Select a main table and then ticked "Partition Table" at
            bottom of wizard screen
            Finish -> Java error & stack trace....

            So I didn't get even as far as you :-(

            Does the above make sense?


            Ellen.

-- [email protected] <mailto:[email protected]>
            JDRF/WT Diabetes and Inflammation Laboratory (DIL)
            Cambridge Institute for Medical Research (CIMR)
            Wellcome Trust/MRC Building
            Addenbrooke's Hospital Hills Road Cambridge CB2 0XY
            Tel:01223 763215  <tel:01223%20763215>

            On 12/12/12 10:30, Arek Kasprzyk wrote:
            Hi Ellen,
            I have now had an opportunity to try your file and it
            seemed to have produced the correct config structure (if
            I understood your problem correctly)

            This is what I did:

            1. Imported your schema
            2. Added data source with one main table
            3. Invoked 'data source management' window
            4. Added second dataset
            5. Updated both datasets
            6. Created a config
            7. Started the server

            I was able to query it through GUI (did not get any
            results bc the schema was empty but did not get any
            errors either)


            a







            On Thu, Dec 6, 2012 at 2:19 PM, Ellen Schofield
            <[email protected]
            <mailto:[email protected]>> wrote:

                Hi Arek,

                Sent to you directly.

                Many thanks,

                Ellen.

-- [email protected] <mailto:[email protected]>
                JDRF/WT Diabetes and Inflammation Laboratory (DIL)
                Cambridge Institute for Medical Research (CIMR)
                Wellcome Trust/MRC Building
                Addenbrooke's Hospital Hills Road Cambridge CB2 0XY
                Tel:01223 763215  <tel:01223%20763215>

                On 06/12/12 14:10, Arek Kasprzyk wrote:
                Hi Ellen,
                would you mind sending me your schema so I can
                double check here?


                cheers,
                a



                On Thu, Dec 6, 2012 at 10:25 AM, Ellen Schofield
                <[email protected]
                <mailto:[email protected]>> wrote:

                    Hi All,

                    I hope somebody is able to help me as I'm
                    struggling with this.  I am using BioMart
                    0.8rc6 and MartConfigurator as part of that.  I
                    have a db with 5 different datasets, all of
                    which have the same structure (one main table
                    and one dm table).  I can use MartConfigurator
                    with one dataset to create a working mart for
                    the main/dm table in that dataset, suggesting
                    that the structure is correct.  However, if I
                    try building a mart on one of the main tables
                    (with it's dimension table) and request
                    building it with a partition table causes a
                    java null pointer exception
                    (MartController.java line 1315).  Building the
                    same partition-based mart without the dm tables
                    present works.

                    Am really hoping somebody can point out what
                    I'm doing wrong?

                    Many thanks.
                    Ellen.

-- [email protected]
                    <mailto:[email protected]>
                    JDRF/WT Diabetes and Inflammation Laboratory (DIL)
                    Cambridge Institute for Medical Research (CIMR)
                    Wellcome Trust/MRC Building
                    Addenbrooke's Hospital Hills Road Cambridge CB2 0XY
                    Tel: 01223 763215 <tel:01223%20763215>

                    _______________________________________________
                    Users mailing list
                    [email protected] <mailto:[email protected]>
                    https://lists.biomart.org/mailman/listinfo/users










_______________________________________________
Users mailing list
[email protected]
https://lists.biomart.org/mailman/listinfo/users

Reply via email to