Discussion List Archives

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: New dictionary: multiblock CIF

Dear James,
Thanks for the example. As I thought about how multiblocks could beimplemented, I got more questions.
Values of column A (using the example from Fig. 3.1.4.1.) can bedropped – do I understand correctly that dropping is optional?
If we had A with four values, we could have four values in one block,or four blocks with one value each. Could we also have 1+3 or 2+2?
In the article I read:> when all the values of a data name in a table are iden-> tical, that data item can be provided instead as a single> value outside the table but within the containerI'm more familiar with CIF1 and DDL2, so I wondered, is it true ingeneral in CIF2.0/DDLm?
Best wishesMarcin
On Fri, Sep 29, 2023 at 10:12 AM James H <jamesrhester@gmail.com> wrote:>> Dear Marcin,>> There are no fully-fledged examples as yet, as this is building the infrastructure for them. The first proper examples, when they appear, are likely to be from powder diffraction, where multiple phases and measurements are quite common. Nevertheless, a toy example is available here: https://github.com/COMCIFS/cif_core/blob/master/examples/cell-measurement-multi-block.cif . Part of the motivation for this particular example was needing to retain the ability for the cell measurement diffraction conditions to be different to the main data set measurement conditions, because in the very old days that could actually happen. This is achieved by putting the cell measurement diffraction conditions in a separate data block, and referring to the _diffrn.id of that data block in _cell_measurement.condition_id.>> Most of the data names in the new multi-block dictionary are used in that example.>> Future development of the multi-block dictionary will expand the list of categories that can be spread over multiple blocks, so, e.g. you will be able to refer to a structure that is described in a different data block, which is useful when capturing parent-child structure relationships amongst other things. If you (or anyone else) is interested in this work you are welcome to follow along and contribute at the Github repository.>> all the best,> James.>> On Fri, 29 Sept 2023 at 01:06, Marcin Wojdyr <wojdyr@gmail.com> wrote:>>>> Dear James,>>>> Thanks for announcing this interesting extension. The simplified>> example in Fig. 3.1.4.1 is a big help in understanding what the>> concepts described in the paper mean. Do you have a full real-world>> example of a CIF file in the new multi-block scheme?>>>> Thanks,>> Marcin>>>> On Thu, Sep 28, 2023 at 7:26 AM James H via cif-developers>> <cif-developers@iucr.org> wrote:>> >>> > Dear COMCIFS, core CIF DMG and cif-developers:>> >>> > Apologies for cross-posting, but I think this is relevant to you all.>> >>> > We have decided to create a new dictionary that contains the additional definitions required when describing data that are distributed over more than one data block. This allows core CIF to remain free of the many "boilerplate" definitions that are required. A draft first release of this multiblock dictionary is available at https://github.com/COMCIFS/MultiBlock_Dictionary/blob/rc-1.0.0/multi_block_core.dic for your review. This first release of the dictionary contains only definitions that are already present in the latest release of the core dictionary; once the present dictionary is approved they will be removed from the core dictionary.>> >>> > The technical details of describing multi-block data with CIF dictionaries are available in section 3.1.4 of our IUCr2023 workshop handout at https://www.iucr.org/__data/assets/pdf_file/0015/157011/handout.pdf, or a brief summary is in the README of the above repository.>> >>> > Please feel free to provide feedback either via reply to this message or raising issues in the Github repository linked above. The review period is one month.>> >>> > best wishes,>> > James.>> > -->> > T +61 (02) 9717 9907>> > F +61 (02) 9717 3145>> > M +61 (04) 0249 4148>> > _______________________________________________>> > cif-developers mailing list>> > cif-developers@iucr.org>> > http://mailman.iucr.org/cgi-bin/mailman/listinfo/cif-developers>>>> --> T +61 (02) 9717 9907> F +61 (02) 9717 3145> M +61 (04) 0249 4148_______________________________________________cif-developers mailing listcif-developers@iucr.orghttp://mailman.iucr.org/cgi-bin/mailman/listinfo/cif-developers

Reply to: [list | sender only]