/tblutils

Primary LanguagePython

Table Utilities ~ tblutils

tblutils is a package for working with tables in Python. It provides a variety of utilities to make working with tables easier.

Table data is flat, with columns and rows, like in a relational database.

Getting Started

tblutils consists of two classes:

  • Column - The basic data type for series (e.g. Age: [45, 22, 30], Favorite Color: ['red','green','blue'])
  • Table - A data type containing one or more columns (e.g. 3 rows <> [Age, Favorite Color])

Loading Data

Tables load data through the load() method, which is invoked upon instantiating the class. For example, if data is in a csv file named "data.csv", it can be loaded during instantiation like:

tbl = Table("data.csv")

or afterwards like:

tbl = Table()
tbl.load("data.csv")

In addition to the data parameter, __init__() and load() take two other parameters:

load(self, data, data_type=None, headers=True)
__init__(self, data=None, data_type=None, headers=True)
  • data_type Str::None - currently supported data types are: 'csv' and 'json' for .csv & .json files, and 'list','dict', 'column' for lists of: lists, dictionaries, columns. If data_type is None, load() invokes the _get_data_type() method to guess the data type.
  • headers Bool::True - flag to indicate if headers are included in the dataset (only applicable for .csv and list-of-list data types)

Saving Data

Currently, the Table class supports writing data in .csv and .json formats. The correct format will be detected by the file name you provide via the _get_data_type() method.

tbl = Table("data.csv")

# to save to a .csv file named 'output.csv'
tbl.save("output.csv")

# to save to a .json file named 'output.json'
tbl.save("output.json")

One can immediately see the ease of converting file types.

Accessing and Displaying Data

The data() method will return the serialized contents of the table. It takes one parameter, _format='dict', to indicate whether to return the data as a list of list or a list of dictionaries.

x = Column([1,5,9,15])
y = Column([1,2,3,4])
tbl = Table([x,y])
print(table.data()) # defaults to list of dictionaries

###
# [{'Column1': 1, 'Column2': 1}, {'Column1': 5, 'Column2': 2}, {'Column1': 9, 'Column2': 3}, {'Column1': 15, 'Column2': 4}]
###

print(tbl.data('list')) # list of lists

###
# [['Column1', 'Column2'], [1, 1], [5, 2], [9, 3], [15, 4]]
###

Columns within a table can be accessed in two ways:

  1. The first is to use their integer index and standard slice notation (e.g. tbl[0] # returns first column). Note, this means that enumerate(tbl) is a column-wise generator, row-wise enumeration can be achieved via enumerate(tbl.data()).
  2. Secondly, they can be accessed by their .label (column name) through the table's select() method (e.g. tbl.select('Column1')). Multiple labels can be passed to select() and a new table is returned containing only the specified columns.