eosc-cos4cloud / mecoda-orange Goto Github PK

Relax strict check of ICTIO_BASIC_XXXXXXXX.zip for ZIP file name

Some of the "features" (columns) of Ictio's database shouldn't be treated as categorical type, since they consume too much memory and Orange tends to "freeze". After testing, we found out that Orange does not actually freeze, it just takes a lot of time minutes to process all data, and meanwhile it's not responsive, specially in Windows.

Instead, the problematic columns should be converted from categorical type to text type (that is, treated as metadata) in order to avoid resource issues. Based on Orange's autodetection capabilities, the columns that should be considered for being treated as metadata are:

• obd_id
• taxon_code
• scientific_name
• number_of_fish
• obs_comments
• user_id
• checklist_id
• port
• location_type
• watershed_code
• watershed_name

Add date filters in ICTIO widget, similar to the ones implemented in the widgets for other Citizen Observatories like OdourCollect:

Enable the ICTIO widget to allow direct load load of the XLSX file containing a valid ICTIO database (typically, the BDB_xxxxxx.xlsx that can be found inside the ICTIO_BASIC_XXXXXXXX.zip file).

Don't rewrite code. You don't need to, since you can still rely on IctioPy lib for that task. In order to load an ICTIO XLSX file instead of an ICTIO ZIP, do the following:

dataframe = ictiopy.sanitizedb(ictiopy.load_ictio_bdb_file(path_to_folder_with_xlsx_file, name_of_xlsx_file_with_extension))

Using this code, the output will be exactly the same as usual.

After that, some help text messages will have to be modified. We can add this text just above the "NOTE":

You can also specity an XLSX file if you have already extracted the file BDB_XXXXXXXX.xlsx that can be found inside ICTIO_BASIC_XXXXXXXX.zip

Hi there, while teaching I had problems when I was downloading the data with my students.

We all downloaded this Monday and Tuesday (morning in Europe) data from the kit 15822 data from January for each second.

Some of them got all the data (around 122 k rows) and some of them they had less data (around 1 500 rows).

Make this dialog resizable for the information message to be more readable (or move info message to a label inside the dialog)