This might be relevant or useful to implement at the same time as #56 and #44 . This would allow the mass processing of data and would be nice to retain memory from previous import attempts to enable repeated analysis as data is reploted/checked during the cycling process.
This might be relevant or useful to implement at the same time as #56 and #44 . This would allow the mass processing of data and would be nice to retain memory from previous import attempts to enable repeated analysis as data is reploted/checked during the cycling process.