This is a pre-production deployment of Warehouse. Changes made here affect the production instance of PyPI (pypi.python.org).
Help us improve Python packaging - Donate today!

Active learning models

Project Description
`dossier.models` is a Python package that provides experimental active learning
models. They are meant to be used as search engines through `dossier.web` web
services.

### Installation

`dossier.models` is on PyPI and can be installed with `pip`:

```bash
pip install dossier.models
```

Currently, `dossier.models` requires Python 2.7. It is not yet Python 3
compatible.


### Documentation

API documentation with examples is available as part of the Dossier Stack
documentation:
[http://dossier-stack.readthedocs.org](http://dossier-stack.readthedocs.org#module-dossier.models)


### Running a simple example

`dossier.models` comes with an example web application that demonstrates how to
use all of the Dossier Stack components to do active learning. The following is
a step-by-step guide to get you up and running with a simple example of
SortingDesk. This guide assumes basic familiarity with standard Python tools
like `pip` and `virtualenv`.

This guide also requires a database of some sort to store data. You can use
any of the backends supported by [kvlayer](https://github.com/diffeo/kvlayer)
(like PostgreSQL, HBase or MySQL). For this guide, we'll use Redis since it
requires very little setup. Just make sure it is installed and running on your
system.

Here are a couple of screenshots of SortingDesk in action:

[![SortingDesk at rest](http://i.imgur.com/I0qT4M9s.png)](http://i.imgur.com/I0qT4M9.png)
[![SortingDesk drag & drop](http://i.imgur.com/Uxeksx5s.png)](http://i.imgur.com/Uxeksx5.png)

First, you should create a new Python virtual environment and install
`dossier.models` from PyPI:

```bash
$ virtualenv dossier
$ source ./dossier/bin/activate
$ pip install dossier.models
```

Depending upon your system setup, this may take a bit of time since
`dossier.models` depends on `numpy`, `scipy` and `scikit-learn`.

Now verify that `dossier.models` is installed correctly:

```bash
$ python -c 'import dossier.models'
```

If all is well, then the command should complete successfully without any
output.

Next, we need to setup configuration so that Dossier Stack knows which database
to use and which indexes to create on feature collections. You can grab
a sample configuration from GitHub:

```bash
$ curl -O https://raw.githubusercontent.com/dossier/dossier.models/master/data/config.yaml
```

The config looks like this:

```yaml
kvlayer:
app_name: dossier
namespace: models
storage_type: redis
storage_addresses: ['localhost:6379']

dossier.store:
feature_indexes: ['name', 'keywords']
```

The first section configures your database credentials. This config assumes
you're using Redis running on `localhost` on port `6379` (the default).

The second section tells Dossier Stack which indexes to create on feature
collections. This configuration is dependent on the features in your data.
In this sample configuration, we've chosen `name` and `keywords` because both
are features in the sample data set.

To download and load the sample data set, grab it from GitHub and use the
`dossier.store` command to load it:

```bash
$ curl -O https://raw.githubusercontent.com/dossier/dossier.models/master/data/example.fc
$ dossier.store -c config.yaml load --id-feature content_id example.fc
```

The `dossier.store` command allows you to interact with feature collections
stored in your database. The `--id-feature` flag tells `dossier.store` to use
the value of the `content_id` feature as the feature collection's primary key.
If this flag is omitted, then a `uuid` is generated instead.

You can confirm that data was added to your database with the `ids` command:

```bash
$ dossier.store -c config.yaml ids
doc11
doc12
doc21
doc22
doc23
...
```

Finally, you can run the web application bundled with `dossier.models`:

```bash
$ dossier.models -c config.yaml
```

Open your browser to
[http://localhost:8080/SortingDesk](http://localhost:8080/SortingDesk) to
see an example of `SortingDesk` with the sample data. If you click on the `X`
link on an item in the queue, a negative label will be added between it and the
query indicated at the top of the page. Or you can drag an item from the queue
into a bin---or drop it anywhere on the body page to create a new bin. Bins can
also be dragged on to other bins to merge them. Go ahead and try it. You can
confirm that a label was made with the `dossier.label` command:

```bash
$ dossier.label -c config.yaml list
Label(doc22, doc42, annotator=unknown, 2014-11-26 16:02:01, value=CorefValue.Negative)
```

You should also be able to see labels being added in the output of the
`dossier.models` command if you're running it in your terminal.

There is also a simpler example using plain `SortingQueue` available at
[http://localhost:8080/SortingQueue](http://localhost:8080/SortingQueue).
Release History

Release History

History Node

0.6.17.dev62

History Node

0.6.17.dev61

History Node

0.6.17.dev55

History Node

0.6.17.dev50

History Node

0.6.17.dev49

History Node

0.6.17.dev48

History Node

0.6.17.dev47

History Node

0.6.17.dev46

History Node

0.6.17.dev45

History Node

0.6.17.dev44

History Node

0.6.17.dev42

History Node

0.6.17.dev40

History Node

0.6.17.dev39

History Node

0.6.17.dev38

History Node

0.6.17.dev37

History Node

0.6.17.dev36

History Node

0.6.17.dev35

History Node

0.6.17.dev34

History Node

0.6.17.dev33

History Node

0.6.17.dev32

History Node

0.6.17.dev30

History Node

0.6.17.dev29

History Node

0.6.17.dev19

History Node

0.6.17.dev17

History Node

0.6.17.dev11

History Node

0.6.17.dev9

History Node

0.6.17.dev8

History Node

0.6.17.dev6

History Node

0.6.17.dev3

This version
History Node

0.6.16

History Node

0.6.16.dev5

History Node

0.6.15

History Node

0.6.15.dev3

History Node

0.6.15.dev2

History Node

0.6.15.dev1

History Node

0.6.14

History Node

0.6.13

History Node

0.6.13.dev7

History Node

0.6.13.dev6

History Node

0.6.13.dev5

History Node

0.6.13.dev3

History Node

0.6.13.dev2

History Node

0.6.13.dev1

History Node

0.6.12

History Node

0.6.12.dev1

History Node

0.6.11

History Node

0.6.11.dev10

History Node

0.6.11.dev9

History Node

0.6.11.dev8

History Node

0.6.11.dev7

History Node

0.6.11.dev1

History Node

0.6.10

History Node

0.6.10.dev1

History Node

0.6.9

History Node

0.6.8

History Node

0.6.7

History Node

0.6.7.dev5

History Node

0.6.7.dev4

History Node

0.6.6

History Node

0.6.6.dev1

History Node

0.6.5

History Node

0.6.4

History Node

0.6.3

History Node

0.6.3.dev18

History Node

0.6.3.dev13

History Node

0.6.3.dev12

History Node

0.6.3.dev9

History Node

0.6.3.dev8

History Node

0.6.3.dev7

History Node

0.6.3.dev6

History Node

0.6.3.dev1

History Node

0.6.2

History Node

0.6.2.dev79

History Node

0.6.2.dev77

History Node

0.6.2.dev76

History Node

0.6.2.dev74

History Node

0.6.2.dev73

History Node

0.6.2.dev72

History Node

0.6.2.dev67

History Node

0.6.2.dev55

History Node

0.6.2.dev54

History Node

0.6.2.dev52

History Node

0.6.2.dev50

History Node

0.6.2.dev47

History Node

0.6.2.dev39

History Node

0.6.2.dev38

History Node

0.6.2.dev36

History Node

0.6.2.dev35

History Node

0.6.2.dev32

History Node

0.6.2.dev23

History Node

0.6.2.dev21

History Node

0.6.2.dev19

History Node

0.6.2.dev18

History Node

0.6.2.dev16

History Node

0.6.2.dev15

History Node

0.6.2.dev14

History Node

0.6.2.dev12

History Node

0.6.2.dev10

History Node

0.6.2.dev9

History Node

0.6.2.dev8

History Node

0.6.2.dev6

History Node

0.6.2.dev3

History Node

0.6.2.dev2

History Node

0.6.2.dev1

History Node

0.6.1

History Node

0.6.1.dev1

History Node

0.6.0

History Node

0.5.1.dev11

History Node

0.5.1.dev9

History Node

0.5.1.dev4

History Node

0.5.1.dev2

History Node

0.5.1.dev1

History Node

0.5.0

History Node

0.3.1.dev14

History Node

0.3.1.dev13

History Node

0.3.1.dev12

History Node

0.3.1.dev11

History Node

0.3.1.dev8

History Node

0.3.1.dev6

History Node

0.3.1.dev4

History Node

0.0.0

Download Files

Download Files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

File Name & Checksum SHA256 Checksum Help Version File Type Upload Date
dossier.models-0.6.16-py2.7.egg (189.7 kB) Copy SHA256 Checksum SHA256 2.7 Egg Jul 8, 2015
dossier.models-0.6.16.tar.gz (99.5 kB) Copy SHA256 Checksum SHA256 Source Jul 8, 2015

Supported By

WebFaction WebFaction Technical Writing Elastic Elastic Search Pingdom Pingdom Monitoring Dyn Dyn DNS Sentry Sentry Error Logging CloudAMQP CloudAMQP RabbitMQ Heroku Heroku PaaS Kabu Creative Kabu Creative UX & Design Fastly Fastly CDN DigiCert DigiCert EV Certificate Rackspace Rackspace Cloud Servers DreamHost DreamHost Log Hosting