This repo presents an idea of fields management from/to different data sources.

Requirements 📐

• Python 3.8.x (Tested in Python 3.8.5)
• Python Pandas

Target Audience 📢

• Data Engineers
• Data Scientists
• Data Analysts
• Software Engineers

Problem 😕

A very common headache in data processing from multiple sources is the managment of column names. In a nutshell, few problems are:

• Lack of naming convention: Column names are not following the same convention or any convention at all.
• Unefficient refactoring: Hardcoded strings are all over the project. In refactoring, its possible to forgot to update something.
• Column dublications: Column names are not in the same place, so its possible the user will use similar column names for the same thing.

Solution 👌

Organize all the fields (column names) in a cetral place by using a Python object (class). The benefits from this strategy is the elimination of the above problems and much more (e.g adds fields logic)!! A Python class gives the ability to use methods and also bound other metadata with fields (e.g type).

An example 😋

Below there is a simple example, but you can find more code examples in examples.ipynb Notebook

Let's say that we have a pandas dataframe with one column.

input_df = pd.DataFrame({'dummy_column_1': [1,2,3,4]})

Output:

+----+----------------+
|    | dummy_column_1 |
|----+----------------|
|  0 |              1 |
|  1 |              2 |
|  2 |              3 |
|  3 |              4 |
+----+----------------+

Now, we want to create a new column dummy_column_2. The first step, is to create a new entity in Field class. For example:

dummy_field_2 = field(input_name='dummy_column_2', exported_name='Dummy Column 2')

Then we can use the new field in our dataframe.

import Field
df[Fields.dummy_field_2.input_name] = 2

Output:

+----+----------------+------------------+
|    | dummy_column_1 |   dummy_column_2 |
|----+----------------+------------------|
|  0 |              1 |                2 |
|  1 |              2 |                2 |
|  2 |              3 |                2 |
|  3 |              4 |                2 |
+----+----------------+------------------+

Then at the end, after we have finished with the processing steps, we can rename the fields before we export the dataframe by leveraging field functions.

field_renames = Field.get_renames()
df = df.rename(columns={field: field_renames[field] for field in input_df.columns})

Output:

+----+----------------+------------------+
|    |   Dummy Column |   Dummy Column 2 | <---- New column names
|----+----------------+------------------|
|  0 |              1 |                2 |
|  1 |              2 |                2 |
|  2 |              3 |                2 |
|  3 |              4 |                2 |
+----+----------------+------------------+

---

Contributing

• Pull Requests (PRs) are welcome ☺️!

Thanks!

Thanks for your time! This repo is just an idea, the above logic can be expanded and change the way you handle multiple data sources! Cheers 🍻