You may often see df appearing on Python-based data science resources and literature. It is a conventional way to denote the pandas DataFrame structure. pandas lets us perform the otherwise tedious operations on tables (data frames) with simple commands, such as dropna(), merge(), pivot(), and set_index().
pandas is designed to streamline handling processes of common data types, such as time series. While NumPy is more specialized in mathematical calculations, pandas has built-in string manipulation functions and also allows custom functions to be applied to each cell via apply().
Before use, we import the module with the conventional shorthand as:
pd.DataFrame(my_list_or_array)
To read data from existing files, just use the following:
pd.read_csv()
For tab-delimited files, just add ' ' as the separator:
pd.read_csv(sep=' ')
pandas supports data import from a wide range of common file structures for data handling and processing, from pd.read_xlsx() for Excel and pd.read_sql_query() for SQL databases to the more recently popular JSON, HDF5, and Google BigQuery.
pandas provides a collection of handy operations for data manipulation and is considered a must-have in a Python data scientist's or developer's toolbox.
To fully understand and utilize the functionalities, you may want to read more from the official documentation: