pystocktwits-data-utils

Data Tools for the JSON outputs from the pystocktwits wrapper to create datasets. This was made separate from the main repo because only includes wrappers for the web calls to stocktwits and it can be expanded on for the oauth tools.

Link to main repo: https://github.com/khmurakami/pystocktwits

Requirements

Also listed in requirements.txt:

• requests
• texblob
• wordcloud

Install

Install Locally

This has been tested with Python 2.7, 3.3, 3.4, 3.5, 3.6 and 3.6-dev

$ git clone https://github.com/khmurakami/pystocktwits_data_utils
$ cd pystocktwits_data_utils
$ python setup.py install
$ cd pystocktwits
$ python setup.py install

Install inside a Virtualenv

$ git clone https://github.com/khmurakami/pystocktwits_data_utils.git
$ git pull --recurse-submodules
$ cd pystocktwits_data_utils
$ python setup.py install
$ cd pystocktwtis
$ python setup.py install

Update Submodule

$ git submodule update --remote --merge

Using this Library

Example of getting all of the sentiment and creating a dataframe from it.

from pystocktwits_data_utils import PyStockTwitData

data = PyStockTwitData()

# Get all msgs from this company that is specified
list_of_msgs, list_of_sentiment_json = data.get_all_msgs_with_sentiment_by_symbol_id("AAPL")

# Parse out the Bullish, Bearish, or None Sentiment
list_of_sentiment = data.extract_sentiment_statements(list_of_sentiment_json)

# Create a Dataframe
dataframe = pd.DataFrame(
    {'msg': list_of_msgs,
     'sentiment': list_of_sentiment
    })

# Print to see dataframe and save
print(dataframe)
dataframe.to_csv('pystocktwitsdataset.csv')

Leaving your computer on and collecting data into a CSV

from pystocktwits_data_utils import PyStockTwitData

data = PyStockTwitData()

# This needs to be cleaned up

# Create a CSV named VEEV.csv
# Get the company VEEV
# Get the recent 30 messages
# Wait 600 seconds before checking again
# Set limit messages to 30
data.stocktwit_csv_create("VEEV.csv", "VEEV", 30, 600, limit=30):

Example Result

msg	sentiment
$VEEV that was the 110 bounce I called for but it happened very fast	Bullish
#Update(17),$VEEV Feb-15 105 Calls Up +206%.,since alerted on: Jan 23. Peak 262%	Bullish

Samples

Code samples can be found in example_code/python_scripts JSON sample outputs can be found in example_code/sample_json_output

Testing using Unit Tests

Run test script to test if it works properly

$ python tests/pystocktwits_data_utils_tests.py

TODO

• Make better README.md
• Add timestamp to csv generator
• Add Company ID to csv generator
• Create a realtime stream of data flows

Links/Credits

• Link to main repo: https://github.com/khmurakami/pystocktwits
• IcyFlames CSV Duplicates Removal with no new libraries: https://stackoverflow.com/questions/15741564/removing-duplicate-rows-from-a-csv-file-using-a-python-script
• Git ignore was used from this repo: https://github.com/github/gitignore/blob/master/Python.gitignore
• Stocktwits Official Site: https://stocktwits.com/
• Stocktwits Developer Sign Up: https://api.stocktwits.com/developers/contact
• Stocktwits Official APU Documentation: https://api.stocktwits.com/developers/docs