Metadata-Version: 2.1
Name: lapros
Version: 0.2.0
Summary: lapros data for better AI
Home-page: https://github.com/Tre-Xanh/lapros.py/tree/master/
Author: Vo Chi Cong
Author-email: ccvo@live.jp
License: Apache Software License 2.0
Keywords: AI,data,noise
Classifier: Development Status :: 3 - Alpha
Classifier: Intended Audience :: Developers
Classifier: Natural Language :: English
Classifier: Programming Language :: Python :: 3.8
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: License :: OSI Approved :: Apache Software License
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: pip
Requires-Dist: packaging
Requires-Dist: loguru
Requires-Dist: numpy
Requires-Dist: pandas
Requires-Dist: plum-dispatch
Requires-Dist: scipy
Provides-Extra: dev
Requires-Dist: cleanlab ; extra == 'dev'
Requires-Dist: pyarrow ; extra == 'dev'
Requires-Dist: scikit-learn ; extra == 'dev'

LaPros
================

<!-- WARNING: THIS FILE WAS AUTOGENERATED! DO NOT EDIT! -->

## Install

`pip install -U lapros`

## How to use

LaPros works with classifiers. It ranks the suspicious labels given
probabilies by some classification model. You can use normal Python
lists, Numpy arrays or Pandas data. Return values are in a Numpy array
or a Pandas series, the larger the value, the more suspicious are the
coresponding labels.

``` python
from lapros import suspect
```

------------------------------------------------------------------------

### suspect

Rank the suspicious labels given probas from a classifier.

Params: accept numpy arrays, pandas dataframes and series, and normal
Python lists.

- probas: shape n x m, probabilites for possible classes.
- labels: shape n x 1, observed class labels

We can use interger, string or even float labels, given that the
probability dataframe’s columns are indexed by the same label set.

Returns: a Pandas DataFrame including 1 index and 2 columns:

- id (int): the index which is the same to the original data row index
- err (float): the magnitude of suspiciousness, valued between \[0, 1\]
- suspected (bool): whether the data row is suspected as having a label
  error.

``` python
labels = [1, 0, 0, 1, 1];
```

``` python
probas = [
    #
    [0.5, 0.6, 0.7, 0.8, 0.9],
    [0.5, 0.4, 0.3, 0.2, 0.1],
];
```

``` python
suspect(
    probas,
    labels=labels,
)
```

<div>
<style scoped>
    .dataframe tbody tr th:only-of-type {
        vertical-align: middle;
    }

    .dataframe tbody tr th {
        vertical-align: top;
    }

    .dataframe thead th {
        text-align: right;
    }
</style>
<table border="1" class="dataframe">
  <thead>
    <tr style="text-align: right;">
      <th></th>
      <th>err</th>
      <th>suspected</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <th>0</th>
      <td>0.4</td>
      <td>False</td>
    </tr>
    <tr>
      <th>1</th>
      <td>0.3</td>
      <td>False</td>
    </tr>
    <tr>
      <th>2</th>
      <td>0.1</td>
      <td>False</td>
    </tr>
    <tr>
      <th>3</th>
      <td>0.7</td>
      <td>False</td>
    </tr>
    <tr>
      <th>4</th>
      <td>0.9</td>
      <td>True</td>
    </tr>
  </tbody>
</table>
</div>
