I recommend using ipython console while playing with finta.

For this example I'll use 5y of AMZN tick data retrieved from NASDAQ site directly.

Open the ipython console.

from finta import TA

import pandas as pd

Load the .csv:

ohlc = pd.read_csv("HistoricalQuotes.csv", index_col="Date", parse_dates=True)

Now we need to make this ohlc comply to standards.

Column names:

ohlc.columns

We need lowercase column names:

ohlc.columns = ['close', 'volume', 'open', 'high', 'low']

As you can see some of the values in the DataFrame have a "$" prefix. Let's see if we can remove that. You may notice that values have "$" prefix, we must remove that before continuing. This small function bellow will do that for us.

def split(dollar: str) -> float:
    return float(dollar.split("$")[1])

Now apply it to each column:

ohlc["close"] = ohlc["close"].apply(split)

ohlc["low"] = ohlc["low"].apply(split)

ohlc["high"] = ohlc["high"].apply(split)

ohlc["open"] = ohlc["open"].apply(split)

Jump right into it to see how easy it is.

TA.RSI(ohlc).tail(10)

Date
2014-12-26    55.099394
2014-12-24    43.666451
2014-12-23    50.085415
2014-12-22    50.594291
2014-12-19    38.730709
2014-12-18    35.584319
2014-12-17    38.632773
2014-12-16    32.701255
2014-12-15    55.449033
2014-12-12    57.338081
Name: RSI, dtype: float64

Those are daily candles with standard RSI-14. How about weekly candles and EMA-5?

Resample the ohlc:

from finta.utils import resample_calendar

finta.utils has a nice utility: "resample_calendar" which will make nice weekly candles in a jiffy.

weekly_ohlc = resample_calendar(ohlc, "7d")

TA.EMA(weekly_ohlc, 5).tail(10)

2019-10-04    1756.299843
2019-10-11    1766.693228
2019-10-18    1771.388819
2019-10-25    1773.145879
2019-11-01    1778.163920
2019-11-08    1770.309280
2019-11-15    1758.442853
2019-11-22    1778.465235
2019-11-29    1765.803490
2019-12-06    1760.108994
Freq: W-FRI, Name: 5 period EMA, dtype: float64

That's it, you now know the basics of finta.