bibsian / database-development Goto Github PK

Instead of having just one raw table we're going to break it up into 5 different tables. One table for biomass, count, density, percent_cover, and individual.

Changing the names of some attribtube

1. samplingprotocol => datatype (levels: count, density, individual, biomass, percent_cover)
2. samplingunits (levels: area's, volume, length)... make it a nested dropdown menu with all possible measurements (both international and non-international???.. americuh).
3. trt_type (levels: precip, consumer, fire, temp, resource, species-removal)
4. Need to spell all columns; abbreviations are confusing.

When using the GUI sometime the sitetable where user input lat/long and descriptions will drop a site level. This usually happens after original site names are modified and the update button has been pushed. Note, only happens in about 1/30 datasets.

Need to decide which columns to prune from the current version of the database (i.e. where the main_table is broken down by site).

Add the ability to reshape data from wide to long (given column names)

Default entry for observational studies was 'habitat'; seems confusing so need to replace with NA or some other category that's going to be in our dictionary for that column.

We need to come up with a protocol for handling individual level data. This includes data on mark-recaptures, distance-sampling, and presence-absence.

Could potentially A) handle this by specifying individual in the samplingunits column or
B) make a different table specifically for these table.

It would be nice to have a feature where you could upload a table or csv file of translations for taxonomic code to genus and species names. Might be a bit complicated... would have to really think about how to do this with enough flexibility to handle varions case where taxa code and latin names vary in what is provided (i.e. key on translates to genus/spp or key translantes to kingdom/phylum/class/order/family etc.).

We need to standardize the vocabulary for the samplingunits (field).

Need to change the file handler class to change the column names of loaded datasets if they are the same names used in the database. This error causes the update_main method in the user facade class to fail (i.e. last step of program fails).

We would like to make the main table reflect information at the level of the project. This would entail a migration where we rename the main table which represents unique site information (other than lat/long) within projects. Then the unique information add the project level would be added with a one to many relationship from the 'new' main_table and a new (created) 'project_table'.