pdf of user guide + pb to import base made from ISIS BASE

User 247a2c5018

07-02-2008 10:02:17

Is it possible to have the pdf of the user guide of instant JCHEM.





I have a base created from ISIS BASE with 3 level of hierarchy.


I have not seen in the user guide that Instant jChem permit levels of hierarchy so I can not import my base from ISIS in a Instant JCHEm.

ChemAxon fa971619eb

07-02-2008 10:41:15

I'm afraid a PDF version of the docs is not available. Only the JavaHelp version inside IJC and the web version that can be found here:


http://www.chemaxon.com/instantjchem/ijc_latest/docs/user/help/htmlfiles/ijcTOC.html





IJC can handle 3 (or more) levels of hierarchy. However the more you have the less usable things become for obvious reasons. I would not recommend going beyond 3 levels. To get the data from ISIS to IJC you would export as an RDF file and import using the the RDF import wizard (Tools->Chemistry->RDF Import).


The process of importing RDF files is quite complex, as there is massive variation found in RDF files and the format is very loosely defined. Not all files will be importable.





Tim

User 247a2c5018

07-02-2008 14:12:57

Is the level of hierarchy of ISIS BASE corresponds to the relationships in IJC?

ChemAxon fa971619eb

07-02-2008 16:48:38

IJC Relationships correspond to database foreign keys. As such they define the relationship between IJC entities (database tables).


A typical IJC Relationship might represent a many-to-one association between two database tables.





Also important is the IJC Data Tree, which is where a hierarchical data structure is defined. The Data Tree is roughly analogous to the ISIS Hview. A Data Tree defines the data hierarchy, using the Relationships that are defined. For instance there is a Vertex for a particular Entity at the root of the tree, and this Vertex can have a child Vertex that corresponds to any other Entity that has a Relationship to the toot Vertex's Entity.





This is explained in more detail in the section of the user guide name 'Relational Data'


http://www.chemaxon.com/instantjchem/ijc_latest/docs/user/help/htmlfiles/ijcTOC.html

User 247a2c5018

08-02-2008 12:27:01

I create a form view and it is saved automatically. But when I close this view, I can not reopen it. Why?

User 247a2c5018

08-02-2008 12:32:24

sorry I find to reopen the view;





I have problem to create the form view when there is two level of hierarchy.


I used the from widget : table. I can see the datas but when there are no datas, I can not insert datas. Why?

ChemAxon fa971619eb

08-02-2008 13:33:36

The easiest way to start with a form for simple master-detail data is to place a WBTable widget for each of the 2 IJC Entities on the form (e.g with master at the top and the detail at the bottom of the form). When you add the WBTable widget you are prompted for which fields to bind to the table. Choose fields from the appropriate entity each time, or select the entity itself for all it fields.





Then go to browse mode of this form and you should see all rows for the master in the top table, and the corresponding rows of the selected master row in the detail table. As you change the selected row in the master table the rows displayed in the detail table will be updated.





If you see data in the master table but nothing in the detail table then your Data Tree or Relationships have not been created correctly.





Once you have this working you can then improve the design of the form - maybe removing the master table and replacing it with widgets for the individual fields.

User 247a2c5018

08-02-2008 14:12:46

Ok I find how to put data in a table widget; As Isis base, we must select the table and choose new row.





I have 2 more questions.


With the form view, I would like to remove alla datas : with the option delete row, i can only remove 1 row...





With I import my database, I would like associate the field name of an other name which will be put in the sdfile=> ex : compound_identifier becomes identifier in the sdile. Is it possible?

ChemAxon fa971619eb

09-02-2008 17:39:56

In form view you can only delete current row.


In grid view you can select multiple rows and delete them. This is your best option.





I don't fully understand your second question. Are you referring to renaming fields during import or export?

User 247a2c5018

11-02-2008 08:33:23

sorry it's during the export

ChemAxon fa971619eb

11-02-2008 08:38:03

Then no, I'm afraid not. The current names of the fields are used as the field names in the exported file. But the field names in IJC can be renamed, so you could change these before starting the export.





Tim

ChemAxon e189db4705

11-02-2008 13:55:38

tdudgeon wrote:
In form view you can only delete current row.


In grid view you can select multiple rows and delete them. This is your best option.
Actually there is one possibility to delete more rows in form:


1. click at current row selection text field in toolbar (between navigation arrows).


2. you can manually edit the selected row and it's allowed to type also interval for multi-row selection. So type for example 1-1000 and press enter


3. most of widgets in form show text "(multiple selection)", which means widgets cannot display data from more rows at once. Only Table can visualize it


4. press delete row action => all selected rows from the interval will be deleted.

User 247a2c5018

12-02-2008 09:14:07

ok that's works!What is the max of rows which could be inserted?

ChemAxon fa971619eb

12-02-2008 13:57:48

I assume you mean deleted?


There is no fundamental limit, but its probably sensible to limit it to deleting a thousand or so rows at a time. Larger number should work, but each deletion may take a little while. For larger amounts memory will become limiting, as the delete operation is undoable so the deleted data needs to be stored.





Tim

User 247a2c5018

12-02-2008 14:04:51

No, I am talking about the max of rows that we could insert in the database.





The problem of no external name during the export like in Isis Base is a problem for me.

ChemAxon fa971619eb

12-02-2008 14:46:00

Fairly large databases can be handled. Up to a million structures would be fine. Some users have gone to several million and claim that performance is fine, but you will see some performance degradation at this level. This is due to all the data being handled in a desktop application, which could be an a relatively low spec machine with limited resource. it is not a limitation of JChem which can scale very well to 10's of millions of structures (e.g. see http://www.chemaxon.com/forum/viewpost12815.html#12815)





The one thing to definitely bear in mind is that extra memory will need to be allocated to IJC if you handle large data sets. As a guideline a million structures would need at least 100MB EXTRA memory.





In future we will be supporting a server based version of IJC which would allow most of the load to be run on a high spec server, making it possible to handle larger datasets on fairly modest PCs.








Tim

User 247a2c5018

13-02-2008 15:33:11

after deleting rows, the index didn't go back to 1.


How could I do this without deleting and recreating the index?

ChemAxon fa971619eb

14-02-2008 10:03:30

Yes, that is how sequences and autoincrement values work in a database. Deleting a row does not make that value available for a a new row.





Potentially there are some approaches that you could use by going in directly to the database and changing things (e.g re-setting the oracle sequence back to 1 if you are using oracle, or using the appropriate ALTER TABLE.. command for Derby or MySQL), but these are not currently supported in IJC.

User 247a2c5018

21-04-2008 07:48:46

When I would like to import V3000 Sdfile, IJC doesn't recognize the field.


For V2000, I have no problem.


Is IJC supporting V3000 format?

User 247a2c5018

21-04-2008 08:06:55

if the Sdfile contains V2000 and V3000, IJC can recognize the fields.


But If the Sdfile contains only V3000, IJC can not recognize the fields.

ChemAxon fa971619eb

21-04-2008 08:20:05

V3000 Molfile format should be supported.


Can you try reading the file into Marvin View to confirm that the file is OK.


If it appears to be valid and gives problems in IJC please could you send us the file (or a small part of it) so that we can examine it.





Thanks





Tim

User 247a2c5018

21-04-2008 08:38:41

see the file :





I have a hierarchy :





Chemical_functions as father and function_name and function_number as son.


When I have 2 entries like :





$DTYPE ROOT:CHEMICAL_FUNCTIONS(1):FUNCTION_NAME


$DATUM H bond acceptor


$DTYPE ROOT:CHEMICAL_FUNCTIONS(1):FUNCTION_NUMBER


$DATUM 4


$DTYPE ROOT:CHEMICAL_FUNCTIONS(2):FUNCTION_NAME


$DATUM H bond donnor


$DTYPE ROOT:CHEMICAL_FUNCTIONS(2):FUNCTION_NUMBER


$DATUM 1








during the import, I have an error :





duplicate columns names : ROOT_CHEMICAL_FUNCTIONS(1)





Why?

User 247a2c5018

21-04-2008 08:41:56

exemple V3000 not recognize by IJC

User 247a2c5018

21-04-2008 08:53:04

Marvin view can open the V3000 file.

ChemAxon fa971619eb

21-04-2008 09:14:53

Thank you for reporting the problem with recognising V3000 format files.


This problem has been identified and corrected. The fix will be present in IJC 2.3 which will be available very soon.





Tim

ChemAxon fa971619eb

21-04-2008 09:21:42

The standard import process is for "flat" data files. The RDF format you mention contains hierarchical data. This can be imported with the standard import, but you will loose the hierarchical nature of the data. The duplicate names is due to the field/column names being too long and needing to be truncated, which results in them becoming duplicates. You can manually fix this be selecting the appropriate field and editing the field/column names.





Alternatively, if you want to retain the hierarchical nature of the data in the RD file then you can use the RDF import process (Tools -> Chemistry -> RDF import). Note that this process is quite a lot more complex than standard import, and due to the poorly defined specification of the RD file format is somewhat dependent of the nature of the RD file. For documentation on the RDF import process please look here:


http://www.chemaxon.com/instantjchem/ijc_latest/docs/user/help/htmlfiles/moving_data/rdf_import.html





Tim

User 247a2c5018

21-04-2008 09:33:33

Yes I also try with :





Tools -> Chemistry -> RDF import





I have no problem with duplicate columns but I lose the form view that I created.


How to keep my form view?

User 247a2c5018

21-04-2008 11:06:07

Is it possible to save a form view like with Isis Base?

ChemAxon fa971619eb

21-04-2008 11:22:55

The RDF import process works differently to standard import.


It is designed to provide a reproducible process, and it removes all artifacts (including forms etc. from earlier attempts to import.


The best approach is to optimise the import process before spending time generating detailed forms.





Tim

ChemAxon fa971619eb

21-04-2008 11:26:04

Quote:
Is it possible to save a form view like with Isis Base?
Form views are automatically saved, so that the are restored to the same state when you re-open.





If you are asking whether a form definition can be exported (e.g. to a file) then the answer is no. The form is inherently coupled to the data tree it belongs to and would have no use outside this data tree.





Tim