Researcher Z records dataset A using sample B at research facility C / instrument D under set of method/conditions/parameters E (sometimes gathered in a paper lab book). Researcher Z copies the data to his/her external hard-drive and takes it back to his/her home institute F. Facility C will take care of the long term storage of the raw dataset, archived in location Y. Researcher Z explores the data in a phase of exploratory analyses, writing a set of macros/scripts G, using software H in different versions (exploratory!) with data outputs J (in different versions, accordingly), generating a set of figures K (in different versions, accordingly). Subset L(K) (often also later versions of it) will be copied into a manuscript document that is produced in various versions with input from authors. A late version of the manuscript is then published as paper M.
PS: In some cases (e.g. due to a doubt in the workflow or a question from a colleague) researcher Z or (rarely, e.g., in case of scientific misconduct is suspected) another person needs to trace back the whole provenance chain from a plot shown in figure L of M, which can be extremely time-consuming.
Researcher Z records dataset A using sample B at research facility C / instrument D under set of method/conditions/parameters E (sometimes gathered in a paper lab book). Researcher Z copies the data to his/her external hard-drive and takes it back to his/her home institute F. Facility C will take care of the long term storage of the raw dataset, archived in location Y. Researcher Z explores the data in a phase of exploratory analyses, writing a set of macros/scripts G, using software H in different versions (exploratory!) with data outputs J (in different versions, accordingly), generating a set of figures K (in different versions, accordingly). Subset L(K) (often also later versions of it) will be copied into a manuscript document that is produced in various versions with input from authors. A late version of the manuscript is then published as paper M.
PS: In some cases (e.g. due to a doubt in the workflow or a question from a colleague) researcher Z or (rarely, e.g., in case of scientific misconduct is suspected) another person needs to trace back the whole provenance chain from a plot shown in figure L of M, which can be extremely time-consuming.