Mar-20-2022, 08:51 PM
(Mar-19-2022, 02:05 AM)Larz60+ Wrote: Data is arranged serially in a CSV file, when I think of database, I think about tables that are arranged in a DBMS such as sqlite, PostGreSQL, Oracle, MySQL, etc. Tables usually are indexed by one or more fields, and there is a query language SQL that allows access to table data in many different ways, including selecting related date from more tables, all in the same query.
You can store data in a CSV file, but to change that data, you usually read the data in sequentially, looking for the data you want to change, and then writing it back out.
Most everything that I need to store I save in json files, and that works in most instances.
If there is a lot of related data, I'll use PostGreSQL.
There are lots of tutorials available on all of the methods mentioned, and many books written on relational databases and other storage methods.
Lots to learn.
But i load the csv file using pandas and in the tutorials they , i think, refer to info that has been loaded into pandas as a database. Is it a bad use of pandas to load the data from a csv file, make modification to it and then save the modified data as a csv file?