Hello.

If you open the source code qnt/optimizer.py, you can find that the optimizer uses these functions for optimization:

def standard_stats_function(data, output): """ Calculates statistics for the iteration output. :param data: market data :param output: weights :return: dict """ start_date = qns.get_default_is_start_date_for_type(data.name) stat = qns.calc_stat(data, output.sel(time=slice(start_date, None))) return stat.isel(time=-1).to_pandas().to_dict() def standard_stats_to_weight(stat): """ Converts the statistics to weight. :param stat: dict :return: """ res = stat.get('sharpe_ratio', float('-inf')) if math.isfinite(res): return res else: return float('-inf')

You can use other functions if you want. Just specify stats_function and stats_to_weight when you call optimize_strategy.

It is ok that you use your own optimization algorithm. This one is the simplest implementation and it may not suit you perfectly. You can just reuse some code from optimizer.py which relates to workers. It may be useful if you want to speed up your optimizer.

Regards.

@support Thanks for the detailed answer, that seems to be it, here is the final code:

import xarray as xr import qnt.stats as qns import qnt.output as qnout import qnt.data as qndata # single-stock trading data = qndata.futures.load_data(min_date="2005-01-01", assets=["F_ES"]) # attempting an optimal (unrealistic) long-only strategy # by looking at future prices, and investing only if there will be profit next_price_open = data.sel(field="open").shift(time=-1) next2_price_open = data.sel(field="open").shift(time=-2) weights = xr.where(next_price_open < next2_price_open, 1.0, 0.0) # sell short when optimal: # weights = xr.where(next_price_open > next2_price_open, -1.0, weights) weights = qnout.clean(weights, data) qnout.check(weights, data) qnout.write(weights) stats = qns.calc_stat( data, weights, # ignoring slippage for simplicity slippage_factor=0, roll_slippage_factor=0) stats.loc[:, "equity"].plot.step();

@support Got it, thanks a lot!

@magenta-grimer thank you!!!

@support Thank you very much for the clarification, and once again congratulations for the great job you are doing 😉

Hello.

I guess you are using a different python version and it can cause the problem.

There is an example of how I configured my project in pycharm. Try to follow my steps. I hope it will help you.

Start a new project and name it 'qntdev' and select "New Conda environment":

238da5ca-9bf9-40fb-abed-5cd96364e775-image.png

Ensure that "Python version" == 3.7.

If you see
73987d80-ac01-4d2e-ab97-27bb1ae208bf-image.png
then just remove the environment folder (just in case the previous env is broken).

When you push the button "create", pycharm will initialize an empty environment with python=3.7.

Then open the command line and enter commands from the instruction:

conda create -n qntdev quantiacs-source::qnt conda-forge::ta-lib conda-forge::dash=1.18 python=3.7 # answer 'y' conda env config vars set -n qntdev API_KEY={your_api_key_here}

Next, close the command line and restart pycharm.

When pycharm finishes indexing, create a strategy.py, put the example code from the documentation to this file. Try to run it (right-click on the file -> 'run').

On the first run you will see the error:

Please, specify the API_KEY.

It looks that pycharm don't see the env variable API_KEY, so just add these 2 lines to the head of the strategy:

import os os.environ['API_KEY'] = '{your_api_key_here}'

After that, start the file again.

f8fa3f12-d226-4bfa-86f6-ccad8185c0b3-image.png

For me, it is working. I think it will suit you too.

Regards.

@damnedlies Hello:

we are taking relative daily returns, in other words the ratio: [price(t)-price(t-1)]/price(t-1)

we are using no benchmark.

Note also that we apply reinvesting when computing the numerator of the Sharpe ratio (geometric mean)

@support said in Local development import error:

conda install -y "quantiacs-source::qnt >= 0.0.225"

Now it works. Thanks!

@rezhak21 Rules are defined at: https://quantiacs.com/contest and more details for the current contests (submission time till end of May) can be found at: https://quantiacs.com/contest/15

For Futures the in sample period starts on January 1st 2006, for the BTC Futures on January 1st, 2014

@anthony_m we patched with spot BTC data see answer: https://quantiacs.com/community/topic/6/btc-contest-start-date

@support understood thnx

@vyacheslav_b Hello, we are using a different slippage model. In the old version of Quantiacs slippage is taken to be 5% (HIGH-LOW). In the new version we use 4% (ATR_14), where ATR_14 is the average true range over the last 14 trading days, and therefore it includes day-to-day price moves in addition to the intraday variation.

@nikos84 We did not export the database of users from Quantiacs Legacy to the new version of Quantiacs. Please make a new account.

@nikos84 Good question. We patched the BTC Futures data with the BTC spot data.