97 resample a grid (#98)

* resampling

* testing the resampling

* Resampling in the notebook

* Resampling in the notebook

* Resampling in the notebook

book/docs/notebooks/pairs.ipynb

@@ -4,10 +4,10 @@
    "cell_type": "markdown",
    "id": "8ec59418-20cd-44a9-9b1d-3695324b5e34",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%% md\n"
-    }
+    },
+    "tags": []
    },
    "source": [
     "# Almost pairs trading"
@@ -17,10 +17,10 @@
    "cell_type": "markdown",
    "id": "8ed6af86-8e44-4f05-b5e5-1a4a8a2aa8cf",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%% md\n"
-    }
+    },
+    "tags": []
    },
    "source": [
     "This little exercise goes back to an idea by Stephen Boyd:\n",
@@ -38,10 +38,10 @@
    "execution_count": 1,
    "id": "d6c865f2-13ab-41c1-ba37-a60be923d722",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [],
    "source": [
@@ -62,10 +62,10 @@
    "execution_count": 2,
    "id": "13346284-0fca-46c7-a9b6-548cdf47a6e4",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [
     {
@@ -109,10 +109,10 @@
    "execution_count": 3,
    "id": "32c28168-9dc9-40c8-9482-dbf50c30c945",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [],
    "source": [
@@ -124,10 +124,10 @@
    "execution_count": 4,
    "id": "e85efcd9-f0f5-458c-b417-825c99842c4c",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [
     {
@@ -1756,7 +1756,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.6"
+   "version": "3.9.7"
   }
  },
  "nbformat": 4,

book/docs/notebooks/quantstats.ipynb

@@ -29,10 +29,10 @@
    "execution_count": 1,
    "id": "e82ec25c-fb0f-4b2d-91ff-963e581fab03",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [],
    "source": [
@@ -50,10 +50,10 @@
    "execution_count": 2,
    "id": "9a6949c3-6c2b-4f28-848d-aca31f6fe53a",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [],
    "source": [
@@ -65,10 +65,10 @@
    "execution_count": 3,
    "id": "9950a6b6-f0c6-4394-9b4c-d1e9dfe427fc",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [],
    "source": [
@@ -80,10 +80,10 @@
    "execution_count": 4,
    "id": "6e4831c4-150a-4b42-a650-7ae0693d58d9",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [],
    "source": [
@@ -95,10 +95,10 @@
    "execution_count": 5,
    "id": "354e42be-dcd1-46a6-a3a2-24914cc2de6d",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [],
    "source": [
@@ -112,10 +112,10 @@
    "execution_count": 6,
    "id": "52a1e0bd-cdd5-40a2-a76f-e61306320288",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [],
    "source": [
@@ -127,10 +127,10 @@
    "execution_count": 7,
    "id": "89f696f5-8820-4201-93df-93fc59734e80",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [
     {
@@ -153,10 +153,10 @@
    "execution_count": 8,
    "id": "036a33ea-0edd-4f7c-8653-270886d6781a",
    "metadata": {
-    "tags": [],
     "pycharm": {
      "name": "#%%\n"
-    }
+    },
+    "tags": []
    },
    "outputs": [
     {
@@ -205,7 +205,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.6"
+   "version": "3.9.7"
   }
  },
  "nbformat": 4,

cvx/simulator/grid.py

@@ -0,0 +1,30 @@
+import pandas as pd
+import numpy as np
+
+
+def resample_index(index, rule):
+    """
+    The resample_index function resamples a pandas DatetimeIndex object
+    to a lower frequency using a specified rule.
+
+
+    Note that the function does not modify the input index object,
+    but rather returns a pandas DatetimeIndex
+    """
+    series = pd.Series(index=index, data=index)
+    a = series.resample(rule=rule).first()
+    return pd.DatetimeIndex(a.values)
+
+
+def project_frame_to_grid(frame, grid):
+    """The project_frame_to_grid function projects a pandas
+    DataFrame object onto a given index grid.
+
+    The function returns a new DataFrame that is only updated for times in the grid,
+    otherwise the previous values carry over.
+
+    Note that the function does not modify the input frame object, but rather returns a new object.
+    """
+    sample = pd.DataFrame(index=frame.index, columns=frame.columns, data=np.NaN)
+    sample.loc[grid] = frame.loc[grid]
+    return sample.ffill()

cvx/simulator/portfolio.py

@@ -1,6 +1,9 @@
 from dataclasses import dataclass
+
+import numpy as np
 import pandas as pd
 
+from cvx.simulator.grid import resample_index, project_frame_to_grid
 from cvx.simulator.trading_costs import TradingCostModel
 
 
@@ -181,6 +184,10 @@ def trades_currency(self) -> pd.DataFrame:
         The resulting dataframe will have the same dimensions as the stocks and prices dataframes. """
         return self.trades_stocks * self.prices.ffill()
 
+    @property
+    def turnover(self) -> pd.DataFrame:
+        return self.trades_currency.abs()
+
     @property
     def cash(self) -> pd.Series:
         """ A property that returns a pandas series representing the cash on hand in the portfolio.
@@ -336,3 +343,43 @@ def truncate(self, before=None, after=None):
                                stocks=self.stocks.truncate(before=before, after=after),
                                trading_cost_model=self.trading_cost_model,
                                initial_cash=self.nav.truncate(before=before, after=after).values[0])
+
+    @property
+    def start(self):
+        """first index with a profit that is not zero"""
+        return self.profit.ne(0).idxmax()
+
+    def resample(self, rule, truncate=False):
+        """The resample method resamples an EquityPortfolio object to a new frequency
+        specified by the rule argument.
+        The method returns a new EquityPortfolio object with the resampled stocks, but prices stay the same.
+
+        If the truncate parameter is set to True, the method first trims the original data
+        to start from the beginning of the EquityPortfolio object's timeline using the truncate method.
+        Otherwise, the original EquityPortfolio timescale is used. The start of trading may be missed
+        as the first point may not be in the resampled grid.
+
+        The function uses a utility function resample_index to create a new DatetimeIndex
+        with the specified resampled rule. The new index is used in the project_frame_to_grid
+        function to translate the stocks DataFrame onto the new grid.
+
+        Finally, a new EquityPortfolio object is created with the original prices
+        DataFrame and the resampled stocks DataFrame. The objects trading cost model and initial cash value
+        are also copied into the new object.
+
+        Note that the resample method does not modify the original EquityPortfolio object,
+        but rather returns a new object.
+        """
+        if truncate:
+            portfolio = self.truncate(before=self.start)
+        else:
+            portfolio = self
+
+        grid = resample_index(portfolio.index, rule=rule)
+
+        stocks = project_frame_to_grid(portfolio.stocks, grid=grid)
+
+        return EquityPortfolio(prices=portfolio.prices,
+                               stocks=stocks,
+                               trading_cost_model=self.trading_cost_model,
+                               initial_cash=self.initial_cash)

tests/test_builder.py

@@ -59,6 +59,7 @@ def test_iteration_state(builder):
         pd.testing.assert_series_equal(state.weights, pd.Series(index=builder.assets, data=0.0), check_names=False)
         pd.testing.assert_series_equal(builder[t[-1]], pd.Series(index=builder.assets, data=0.0), check_names=False)
 
+
 def test_build(builder_weights, prices):
     # build the portfolio directly
     portfolio = builder_weights.build()
@@ -72,3 +73,21 @@ def test_build(builder_weights, prices):
 
     # verify both methods give the same result
     pd.testing.assert_series_equal(portfolio.nav, portfolio2.nav)
+
+
+def test_set_weights(prices):
+    b = _builder(prices=prices[["B", "C"]].head(5), initial_cash=50000)
+    for times, state in b:
+        b.set_weights(time=times[-1], weights=pd.Series(index=["B","C"], data=0.5))
+
+    portfolio = b.build()
+    assert portfolio.nav.values[-1] == pytest.approx(49773.093729)
+
+
+def test_set_cashpositions(prices):
+    b = _builder(prices=prices[["B", "C"]].head(5), initial_cash=50000)
+    for times, state in b:
+        b.set_cashposition(time=times[-1], cashposition=pd.Series(index=["B", "C"], data=state.nav / 2))
+
+    portfolio = b.build()
+    assert portfolio.nav.values[-1] == pytest.approx(49773.093729)

tests/test_grid.py

@@ -0,0 +1,51 @@
+import pytest
+import pandas as pd
+import numpy as np
+
+from cvx.simulator.builder import builder
+from cvx.simulator.grid import resample_index, project_frame_to_grid
+
+
+def test_resample_index(prices):
+    x = resample_index(prices.index, rule="M")
+    for t in x:
+        assert t in prices.index
+
+    assert len(x) == 28
+
+    # first day in the middle of a month
+    x = resample_index(prices.index[4:], rule="M")
+    assert len(x) == 28
+    assert x[0] == prices.index[4]
+
+
+def test_project_frame_to_grid(prices):
+    grid = resample_index(prices.index[4:], rule="M")
+    frame = project_frame_to_grid(prices, grid=grid)
+    a = frame.diff().sum(axis=1)
+    assert a.tail(5).sum() == 0.0
+
+
+
+def test_portfolio_resampling(prices):
+    # construct a portfolio with only one asset
+    b = builder(prices[["A"]])
+    assert set(b.assets) == {"A"}
+
+    # compute a grid and rebalance only on those particular days
+    grid = resample_index(prices.index, rule="M")
+
+    # change the position only at days that are in the grid
+    for times, _ in b:
+        if times[-1] in grid:
+            b[times[-1]] = pd.Series({"A": np.random.rand(1)})
+        else:
+            # new position is the old position
+            # This may look ineffective but we could use
+            # trading costs for just holding short positions etc.
+            b[times[-1]] = b[times[-2]]
+
+    portfolio = b.build()
+    print(portfolio.stocks)
+
+    assert portfolio.stocks["A"].tail(10).std() == pytest.approx(0.0, abs=1e-12)