Reversible Data Transforms
Project description
An open source project from Data to AI Lab at MIT.
RDT: Reversible Data Transforms
- License: MIT
- Development Status: Pre-Alpha
- Homepage: https://github.com/sdv-dev/RDT
Overview
RDT is a Python library used to transform data for data science libraries and preserve the transformations in order to revert them as needed.
Install
Requirements
RDT has been developed and tested on Python 3.6, 3.7 and 3.8 on GNU/Linux, macOS and Windows systems.
Also, although it is not strictly required, the usage of a virtualenv is highly recommended in order to avoid interfering with other software installed in the system where RDT is run.
Install with pip
The easiest and recommended way to install RDT is using pip:
pip install rdt
This will pull and install the latest stable release from PyPi.
If you want to install from source or contribute to the project please read the Contributing Guide.
Quickstart
In this short series of tutorials we will guide you through a series of steps that will help you getting started using RDT to transform columns, tables and datasets.
Transforming a column
In this first guide, you will learn how to use RDT in its simplest form, transforming
a single column loaded as a pandas.DataFrame
object.
1. Load the demo data
You can load some demo data using the rdt.get_demo
function, which will return some random
data for you to play with.
from rdt import get_demo
data = get_demo()
This will return a pandas.DataFrame
with 10 rows and 4 columns, one of each data type supported:
0_int 1_float 2_str 3_datetime
0 38.0 46.872441 b 2021-02-10 21:50:00
1 77.0 13.150228 NaN 2021-07-19 21:14:00
2 21.0 NaN b NaT
3 10.0 37.128869 c 2019-10-15 21:39:00
4 91.0 41.341214 a 2020-10-31 11:57:00
5 67.0 92.237335 a NaT
6 NaN 51.598682 NaN 2020-04-01 01:56:00
7 NaN 42.204396 c 2020-03-12 22:12:00
8 68.0 NaN c 2021-02-25 16:04:00
9 7.0 31.542918 a 2020-07-12 03:12:00
Notice how the data is random, so your output might look a bit different. Also notice how RDT introduced some null values randomly.
2. Load the transformer
In this example we will use the datetime column, so let's load a DatetimeTransformer
.
from rdt.transformers import DatetimeTransformer
transformer = DatetimeTransformer()
3. Fit the Transformer
Before being able to transform the data, we need the transformer to learn from it.
We will do this by calling its fit
method passing the column that we want to transform.
transformer.fit(data['3_datetime'])
4. Transform the data
Once the transformer is fitted, we can pass the data again to its transform
method in order
to get the transformed version of the data.
transformed = transformer.transform(data['3_datetime'])
The output will be a numpy.ndarray
with two columns, one with the datetimes transformed
to integer timestamps, and another one indicating with 1s which values were null in the
original data.
array([[1.61299380e+18, 0.00000000e+00],
[1.62672924e+18, 0.00000000e+00],
[1.59919923e+18, 1.00000000e+00],
[1.57117554e+18, 0.00000000e+00],
[1.60414542e+18, 0.00000000e+00],
[1.59919923e+18, 1.00000000e+00],
[1.58570616e+18, 0.00000000e+00],
[1.58405112e+18, 0.00000000e+00],
[1.61426904e+18, 0.00000000e+00],
[1.59452352e+18, 0.00000000e+00]])
5. Revert the column transformation
In order to revert the previous transformation, the transformed data can be passed to
the reverse_transform
method of the transformer:
reversed_data = transformer.reverse_transform(transformed)
The output will be a pandas.Series
containing the reverted values, which should be exactly
like the original ones.
0 2021-02-10 21:50:00
1 2021-07-19 21:14:00
2 NaT
3 2019-10-15 21:39:00
4 2020-10-31 11:57:00
5 NaT
6 2020-04-01 01:56:00
7 2020-03-12 22:12:00
8 2021-02-25 16:04:00
9 2020-07-12 03:12:00
dtype: datetime64[ns]
Transforming a table
Once we know how to transform a single column, we can try to go the next level and transform a table with multiple columns.
1. Load the HyperTransformer
In order to manuipulate a complete table we will need to load a rdt.HyperTransformer
.
from rdt import HyperTransformer
ht = HyperTransformer()
2. Fit the HyperTransformer
Just like the transfomer, the HyperTransformer needs to be fitted before being able to transform data.
This is done by calling its fit
method passing the data
DataFrame.
ht.fit(data)
3. Transform the table data
Once the HyperTransformer is fitted, we can pass the data again to its transform
method in order
to get the transformed version of the data.
transformed = ht.transform(data)
The output, will now be another pandas.DataFrame
with the numerical representation of our
data.
0_int 0_int#1 1_float 1_float#1 2_str 3_datetime 3_datetime#1
0 38.000 0.0 46.872441 0.0 0.70 1.612994e+18 0.0
1 77.000 0.0 13.150228 0.0 0.90 1.626729e+18 0.0
2 21.000 0.0 44.509511 1.0 0.70 1.599199e+18 1.0
3 10.000 0.0 37.128869 0.0 0.15 1.571176e+18 0.0
4 91.000 0.0 41.341214 0.0 0.45 1.604145e+18 0.0
5 67.000 0.0 92.237335 0.0 0.45 1.599199e+18 1.0
6 47.375 1.0 51.598682 0.0 0.90 1.585706e+18 0.0
7 47.375 1.0 42.204396 0.0 0.15 1.584051e+18 0.0
8 68.000 0.0 44.509511 1.0 0.15 1.614269e+18 0.0
9 7.000 0.0 31.542918 0.0 0.45 1.594524e+18 0.0
4. Revert the table transformation
In order to revert the transformation and recover the original data from the transformed one,
we need to call reverse_transform
method of the HyperTransformer
instance passing it the
transformed data.
reversed_data = ht.reverse_transform(transformed)
Which should output, again, a table that looks exactly like the original one.
0_int 1_float 2_str 3_datetime
0 38.0 46.872441 b 2021-02-10 21:50:00
1 77.0 13.150228 NaN 2021-07-19 21:14:00
2 21.0 NaN b NaT
3 10.0 37.128869 c 2019-10-15 21:39:00
4 91.0 41.341214 a 2020-10-31 11:57:00
5 67.0 92.237335 a NaT
6 NaN 51.598682 NaN 2020-04-01 01:56:00
7 NaN 42.204396 c 2020-03-12 22:12:00
8 68.0 NaN c 2021-02-25 16:04:00
9 7.0 31.542918 a 2020-07-12 03:12:00
History
0.2.9 - 2020-11-27
This release fixes a bug that prevented the CategoricalTransformer
from working properly
when being passed data that contained numerical data only, without any strings, but also
contained None
or NaN
values.
Issues closed
- KeyError: nan - CategoricalTransformer fails on numerical + nan data only - Issue #142 by @csala
0.2.8 - 2020-11-20
This release fixes a few minor bugs, including some which prevented RDT from fully working on Windows systems.
Thanks to this fixes, as well as a new testing infrastructure that has been set up, from now on RDT is officially supported on Windows systems, as well as on the Linux and macOS systems which were previously supported.
Issues closed
- TypeError: unsupported operand type(s) for: 'NoneType' and 'int' - Issue #132 by @csala
- Example does not work on Windows - Issue #114 by @csala
- OneHotEncodingTransformer producing all zeros - Issue #135 by @fealho
- OneHotEncodingTransformer support for lists and lists of lists - Issue #137 by @fealho
0.2.7 - 2020-10-16
In this release we drop the support for the now officially dead Python 3.5 and introduce a new feature in the DatetimeTransformer which reduces the dimensionality of the generated numerical values while also ensuring that the reverted datetimes maintain the same level as time unit precision as the original ones.
- Drop Py35 support - Issue #129 by @csala
- Add option to drop constant parts of the datetimes - Issue #130 by @csala
0.2.6 - 2020-10-05
0.2.5 - 2020-09-18
Miunor bugfixing release.
Bugs Fixed
- Handle NaNs in OneHotEncodingTransformer - Issue #118 by @csala
- OneHotEncodingTransformer fails if there is only one category - Issue #119 by @csala
- All NaN column produces NaN values enhancement - Issue #121 by @csala
- Make the CategoricalTransformer learn the column dtype and restore it back - Issue #122 by @csala
0.2.4 - 2020-08-08
General Improvements
0.2.3 - 2020-07-09
- Implement OneHot and Label encoding as transformers - Issue #112 by @csala
0.2.2 - 2020-06-26
Bugs Fixed
- Escape
column_name
in hypertransformer - Issue #110 by @csala
0.2.1 - 2020-01-17
Bugs Fixed
- Boolean Transformer fails to revert when there are NO nulls - Issue #103 by @JDTheRipperPC
0.2.0 - 2019-10-15
This version comes with a brand new API and internal implementation, removing the old
metadata JSON from the user provided arguments, and making each transformer work only
with pandas.Series
of their corresponding data type.
As part of this change, several transformer names have been changed and a new BooleanTransformer and a feature to automatically decide which transformers to use based on dtypes have been added.
Unit test coverage has also been increased to 100%.
Special thanks to @JDTheRipperPC and @csala for the big efforts put in making this release possible.
Issues
- Drop the usage of meta - Issue #72 by @JDTheRipperPC
- Make CatTransformer.probability_map deterministic - Issue #25 by @csala
0.1.3 - 2019-09-24
New Features
- Add attributes NullTransformer and col_meta - Issue #30 by @ManuelAlvarezC
General Improvements
- Integrate with CodeCov - Issue #89 by @csala
- Remake Sphinx Documentation - Issue #96 by @JDTheRipperPC
- Improve README - Issue #92 by @JDTheRipperPC
- Document RELEASE workflow - Issue #93 by @JDTheRipperPC
- Add support to Python 3.7 - Issue #38 by @ManuelAlvarezC
- Create way to pass HyperTransformer table dict - Issue #45 by @ManuelAlvarezC
0.1.2
- Add a numerical transformer for positive numbers.
- Add option to anonymize data on categorical transformer.
- Move the
col_meta
argument from method-level to class-level. - Move the logic for missing values from the transformers into the
HyperTransformer
. - Removed unreacheble lines in
NullTransformer
. Numbertransfomer
to set default value to 0 when the column is null.- Add a CLA for collaborators.
- Refactor performance-wise the transformers.
0.1.1
- Improve handling of NaN in NumberTransformer and CatTransformer.
- Add unittests for HyperTransformer.
- Remove unused methods
get_types
andimpute_table
from HyperTransformer. - Make NumberTransformer enforce dtype int on integer data.
- Make DTTransformer check data format before transforming.
- Add minimal API Reference.
- Merge
rdt.utils
intoHyperTransformer
class.
0.1.0
- First release on PyPI.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.