Lemuras 1.1.7

A small Python library to deal with big tables

Lemuras

Sometimes you cannot or don't want to use Pandas or similar advanced tool for data analysis, but still have a need to manipulate large tables. In such cases you can use Lemuras – it is a pure Python library in a single file without dependencies. And if you have some experience of Pandas or SQL, then you can easily work with Lemuras.

Again, this library may be considered as a simplified analogue of Pandas, but not as a replacement. However, Lemuras is capable of processing an operation on a few tables with several thousands of rows in less than a second on a simple web server.

Features

• Integration with Jupyter IPython Notebook: Lemuras objects are printed as nice tables.
• Most of the syntax is very similar to Pandas.
• Automatic columns types detection.
• Save / load CSV files, JSON, HTML tables, SQL (both query result and table creation code).
• Cells access, rows, columns adding, deleting, columns renaming, functions/lambdas applying, rows sorting.
• Dealing with columns: you can take a table column, do math with the values, compare, check existing in other column or list, then filter a table by it or add it to a table, etc.
• Grouping by none, one, or multiple columns, aggregation with built-in or user-defined functions and lambdas.
• Merge (Join): inner / left / right / outer.
• Tables concatenation and appending.
• Pivot tables creation.

It was tasted on both Python 2.7 and Python 3.6

Examples

All the features are described in notebook examples:

1. Basic things – access to columns, cells, rows; add, delete, change their values; also filtering and sorting.
2. Group by – grouping and combining (aggregating).
3. Merge / Join – such types: inner, outer, left, right.
4. Pivot table – create new tables with columns, rows and cells from another table.
5. Tables Concatenate / Append – simple tables concatenation and appending.
6. Types, Read/Write, CSV, SQL, JSON, HTML – description of Lemuras supported data types, saving to and loading from CSV, SQL, JSON, HTML formats.

In addition, there is one complex example of solving a real-life problem:

• Retargeting results report

The source code of Lemuras is well-commented, so, you can find useful information there. Contributions are welcome.