- OandaHistoricalData.py : is for downloading Forex Data from Oanda, by using Oanda's provided rest api.
Get data in pandas dataframe using sample syntax like:
ohd = OandaHistoricalData()
df = ohd.getCandlesData(insrtument='GBP_USD', count=5, candleFormat='midpoint' ,granularity='H4', start=start_time, end=end_time, csv='Y')
OR
use download_all_data.py to download data from multiple instruments in csv format.
- download_all_data.py : is for downloading candle data of 10 currency insturments for different time-intervals When you run this code, set start_time, end_time, granualarities and instruments, as per your needs.
- start_time = '2000-01-01 00:00'
- end_time = '2018-05-20 22:00'
- granularities = ['M1','M5','M15','M30','H1','H2','H4','H8','D','W','M']
- insrtuments = ['EUR_USD','GBP_USD','USD_CAD','USD_CHF','USD_JPY','EUR_GBP','EUR_CHF','AUD_USD','EUR_JPY','GBP_JPY']
- download_one_instrument.py : is for downloading candle data of 1 currency insturments for specific time-interval. When you run this code, set start_time, end_time, granualarity and instrument, as per your needs.
- start_time = '2017-06-01 00:00'
- end_time = '2017-06-20 22:00'
- granularity = 'H1'
- insrtument = 'EUR_USD'
- WrangleOandaData.py : is for processing downloaded forex candle data, and adding columns like
- Bollinger Band
- High Low difference
- High Open difference
- Low Open difference
- Percentage Change
- Bands of pct change
- Bands of pips change
You can modify this class to add your favorite studies.
- prepare_csv_for_deeplearning.py : is for creating csv files with processed data. It creates two csv files, one is with all columns, and other is with less columns, which are mainly focused on deep learning.
Do not forget to change following paths
- oanda_file_dir = 'exact path where downloaded files are kept'
- processed_oanda_file_dir = 'exact path where new processed csv files can be stored'
- python 2.7
- pandas
- numpy
- requests
- json
- datetime