If a project saved in “Projects” is referencing files in a folder that is accessible from Datagrok, is it possible to configure it to reload the data when the loaded files are updated to the latest version?
As far as I have confirmed, when saving a project data is embedded stably not dynamiccaly, meaning that changes to the loaded files will not be reflected even if they are modified.
Hello, Kosuke Takeuchi,
Thank you for your feedback! To save a project with automatic data updating capability, enable the Data sync option when uploading the project.
In the GIF, I upload the project with Data sync enabled and after that update the dataset. The changes to the dataset automatically update the project. This is noticeable in the scatterplot and the dataframe itself.
Thank you for your quick response and also .gif image. It’s easy to understand, “Seeing is Believing.”
Let me confirm an additional point. In my environment, Data-sync option does not appear. Do have any idea for this phenomenon? I should also ask my colleague in charge about it.
This option appears when you open a dataset from a reproducible source, like querying a database or opening a file from the file share in “Files”. But if you open a local file, this option won’t appear since you cannot “replay” it.
Hi, Andrew,
Thank you for clarification. Unfortunately, I loaded .sdf file from “Files” but data-sync option didn’t show up. Windows file server is used and is conncected to “Files” in my environment. I guess there seems to be an internal configuration issue there, so I will let my collegue in charge know to check.
On the other hand, I am good to hear that I will be able to use data-sync option when loading a synchronizable file once I can fix this issue.
Let me notify that I found that I can use data-sync option when loading .csv file. On the other hand, when loading .sdf file data-sync option disappeared. It would be helpful if I could know whether this issue can be fixed or not. I would appreciate your kind support. Thanks, Kosuke
Thank you for bringing this to our attention. You can monitor the progress of the bug fix through this ticket:
Projects: data-sync option is not shown for .sdf files #2566
Thank you for your help to deal with it.
Hi,
Was it difficult to fix the bug for loading the latest SDF file that I requested the other day? Thank you for considering it.
I have also been considering if there is any alternative solution without using SDF. I checked how to merge compound_structure.csv including id and MolBlock data into the main_table_data.csv file. Although the datasync option appears when using two csv files, an error message, “NullError: method not found: ‘gv’ on null” was shown and I cannot display the merged data when opening the saved project. The behavior of not being able to load the data when using the Joining table may be a new bug. For your information, I could confirm the same error showing when using very simple mock csv files.
Currently, it may be appropriate to prepare a single csv file in advance including table data and structural data as MolBlock format in order to load the latest data without using SDF.
Hi Kosuke,
I’m pleased to inform you that we’ve added the datasync option for SDF files in the 1.17.9 platform version, and you can now save SDF files to the projects with synchronization.
Additionally, I was able to reproduce the error you mentioned, and we will fix it soon. Thank you for the feedback.
Hi,
Thank you for your information!
We are using 1.17.10 platform now, and should we run update command to use datasync for SDF?
And can we apply the option to the project which was made before the update?
Thanks,
Taka
Thank you for your wonderful information. I confirmed datasync behavior when reading SDF file. Great job.
Besides, thank you for checking unwanted behavior about a project using joining table between tables. That would be helpful for someone in future if it can be fixed.
Hi Taka,
There is no need to update 1.17.10 platform to use datasync for SDF. The data synchronization option is automatically available for projects created earlier if the method of obtaining their tables (the .script tag) is saved.
Best regards,
Olena
Hi,
I would like to know current situation aobut unwanted behavior when using joinning table function. It is written just above your reply.
…, an error message, “NullError: method not found: ‘gv’ on null” was shown and I cannot display the merged data when opening the saved project…
Hopefully, this will be fixed soon. If you have any questions or comments, please let me know.
Best regards,
Kosuke
Hi,
Unfortunately, there is no support for cross-table data-sync functions such as join tables.
Datagrok still tries to capture everything that happens with tables and captures joinTable function, but it can’t be reproduced.
We planned to include this support in 1.19.0.
Hi Alex,
Thank you for your sincere effort to tackle this issue.