How to create a strategy with pandas dataframe

[cortezkwan]

Hi there,
If I have a dataframe with multiple futures close prices and the last column is the 'signal' which contains -1, 0 and 1. May I know that what I should do to convert that dataframe to Xarray on def strategy(data)? And what columns are needed, e.g: Assets, signal and time? Thanks

[Vyacheslav_B]

@cortezkwan

Hello

I think these examples can help you.

https://quantiacs.com/documentation/en/examples/quick_start.html#an-example-using-pandas

def get_price_pct_change(prices):
    prices_pandas= prices.to_pandas()
    assets= data.coords["asset"].values
    for asset in assets:
        prices_pandas[asset]= prices_pandas[asset].pct_change()
    return prices_pandas


prices= data.sel(field="close") * 1.0
prices_pct_change= get_price_pct_change(prices).unstack().to_xarray()

How to create a dataset yourself

import pandas as pd
import xarray as xr
import numpy as np


class Fields:
    OPEN = "open"
    LOW = "low"
    HIGH = "high"
    CLOSE = "close"
    VOL = "vol"
    DIVS = "divs"
    SPLIT = "split"
    SPLIT_CUMPROD = "split_cumprod"
    IS_LIQUID = 'is_liquid'


f = Fields


class Dimensions:
    TIME = 'time'
    FIELD = 'field'
    ASSET = 'asset'


ds = Dimensions

dims = (ds.FIELD, ds.TIME, ds.ASSET)


def get_base_df():
    try_result = {"schema": {"fields": [{"name": "time", "type": "datetime"}, {"name": "open", "type": "integer"},
                                        {"name": "close", "type": "integer"}, {"name": "low", "type": "integer"},
                                        {"name": "high", "type": "integer"}, {"name": "vol", "type": "integer"},
                                        {"name": "divs", "type": "integer"}, {"name": "split", "type": "integer"},
                                        {"name": "split_cumprod", "type": "integer"},
                                        {"name": "is_liquid", "type": "integer"}], "primaryKey": ["time"],
                             "pandas_version": "0.20.0"}, "data": [
        {"time": "2021-01-30T00:00:00.000Z", "open": 1, "close": 2, "low": 1, "high": 2, "vol": 1000, "divs": 0,
         "split": 0, "split_cumprod": 0, "is_liquid": 1},
        {"time": "2021-01-31T00:00:00.000Z", "open": 2, "close": 3, "low": 2, "high": 3, "vol": 1000, "divs": 0,
         "split": 0, "split_cumprod": 0, "is_liquid": 1},
        {"time": "2021-02-01T00:00:00.000Z", "open": 3, "close": 4, "low": 3, "high": 4, "vol": 1000, "divs": 0,
         "split": 0, "split_cumprod": 0, "is_liquid": 1},
        {"time": "2021-02-02T00:00:00.000Z", "open": 4, "close": 5, "low": 4, "high": 5, "vol": 1000, "divs": 0,
         "split": 0, "split_cumprod": 0, "is_liquid": 1},
        {"time": "2021-02-03T00:00:00.000Z", "open": 5, "close": 6, "low": 5, "high": 6, "vol": 1000, "divs": 0,
         "split": 0, "split_cumprod": 0, "is_liquid": 1},
        {"time": "2021-02-04T00:00:00.000Z", "open": 6, "close": 7, "low": 6, "high": 7, "vol": 1000, "divs": 0,
         "split": 0, "split_cumprod": 0, "is_liquid": 1},
        {"time": "2021-02-05T00:00:00.000Z", "open": 7, "close": 8, "low": 7, "high": 8, "vol": 1000, "divs": 0,
         "split": 0, "split_cumprod": 0, "is_liquid": 1},
    ]}
    columns = [Dimensions.TIME, f.OPEN, f.CLOSE, f.LOW, f.HIGH, f.VOL, f.DIVS, f.SPLIT, f.SPLIT_CUMPROD,
               f.IS_LIQUID]
    rows = try_result['data']
    for r in rows:
        r['time'] = np.datetime64(r['time'])
    pandas = pd.DataFrame(columns=columns, data=rows)
    pandas.set_index(Dimensions.TIME, inplace=True)
    prices_array = pandas.to_xarray().to_array(Dimensions.FIELD)
    prices_array.name = "BTC_TEST"
    prices_array_r = xr.concat([prices_array], pd.Index(['BTC'], name='asset'))
    return prices_array_r


weights = get_base_df()
print(weights)