Windows installation

• Installation has been tested on Windows 7 and Windows 10
• Fixed (pinned) the versions of some crucial dependencies (tensorflow, keras, numpy)
• 41% doctest coverage

Simpler machine learning example

• diabetes.py and .csv for simplified appendix on machine learning
• more regexes for url recognition/extraction

Manuscript finalized

• loaders for l33t and netspeak dictionary dataset
• python 2.7 tests on travis
• 41% coverage
• miniconda install on travis instead of anaconda
• conda build recipe doesn't work because not all requirements.txt have conda builds available on conda-forge

URL-extracting regex

Add simple URL-extracting regex and import all the more complicated ones from pugnlp:

>>> from nlpia.regexes import RE_URL_SIMPLE
>>> re.findall(RE_URL_SIMPLE, '* Sublime Text 3 (https://www.sublimetext.com/3) is great!')[0][0]
'https://www.sublimetext.com/3'
>>> re.findall(RE_URL_SIMPLE, 'Google github totalgood [github.com/totalgood]!')[0][0]
'github.com/totalgood'

Automagic Downloaders & LSTM for Translation

The following examples from the book should now work:

>>> nlpia.loaders.get_data('deu')  # German-English sentence pairs for `nlpia.translate` LSTM models 
>>> nlpia.loaders.get_data('imdb')  # DataFrame of IMDB movie reviews with ratings

Also, a multi-language international translation character-based LSTM model can be built using the new nlpia.translate module:

>>> from nlpia.translate import *
>>> model = main('spa', n=10000, epochs=100, batch_size=64, num_neurons=128)
Train on 9000 samples, validate on 1000 samples
Epoch 1/100

Also, the base-requirements.txt now includes Keras, tensorflow-gpu, SpaCy, and regex.