wassname/catalyst
1
0
Fork 0
mirror of https://github.com/wassname/catalyst.git synced 2026-06-27 19:30:28 +08:00
Code Issues Packages Projects Releases Wiki Activity

DOC: jupyter notebook in beginner tutorial

Browse Source
This commit is contained in:
Victor Grau Serrat
2017-10-20 14:49:54 -06:00
parent 1b84023c5d
commit 5d5ec6b9be
3 changed files with 241 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files
catalyst/constants.py
+3 -1
View File
@@ -1,3 +1,5 @@
# -*- coding: utf-8 -*-
LOG_LEVEL = 'ERROR'
import logbook
LOG_LEVEL = logbook.INFO
catalyst/examples/buy_btc_simple.py
+3 -3
View File
@@ -1,8 +1,8 @@
from catalyst.api import order, record, symbol
def initialize(context):
context.asset = symbol('btc_usd')
context.asset = symbol('btc_usd')
def handle_data(context, data):
order(asset, 1)
record(btc=data.current(context.asset, 'price'))
order(context.asset, 1)
record(btc = data.current(context.asset, 'price'))
docs/source/beginner-tutorial.rst
+235 -15
View File
@@ -52,7 +52,7 @@ My first algorithm
~~~~~~~~~~~~~~~~~~
Lets take a look at a very simple algorithm from the ``examples``
directory, ``buy_btc.py``:
directory: `buy_btc_simple.py <https://github.com/enigmampc/catalyst/blob/master/catalyst/examples/buy_btc_simple.py>`_:
.. code-block:: python
@@ -225,16 +225,16 @@ Thus, to execute our algorithm from above and save the results to
.. code-block:: python
catalyst run -f buy_btc_simple.py -x bitfinex --start 2016-1-1 --end 2016-9-29 -o buy_simple_btc_out.pickle
catalyst run -f buy_btc_simple.py -x bitfinex --start 2016-1-1 --end 2017-9-30 -o buy_btc_simple_out.pickle
..
.. parsed-literal
.. parsed-literal::
.. AAPL
.. [2015-11-04 22:45:32.820166] INFO: Performance: Simulated 3521 trading days out of 3521.
.. [2015-11-04 22:45:32.820314] INFO: Performance: first open: 2000-01-03 14:31:00+00:00
.. [2015-11-04 22:45:32.820401] INFO: Performance: last close: 2013-12-31 21:00:00+00:00
INFO: run_algo: running algo in backtest mode
INFO: exchange_algorithm: initialized trading algorithm in backtest mode
INFO: Performance: Simulated 639 trading days out of 639.
INFO: Performance: first open: 2016-01-01 00:00:00+00:00
INFO: Performance: last close: 2017-09-30 23:59:00+00:00
``run`` first calls the ``initialize()`` function, and then
@@ -255,7 +255,7 @@ slippage model that ``catalyst`` uses).
Let's take a quick look at the performance ``DataFrame``. For this, we
use ``pandas`` from inside the IPython Notebook and print the first ten
rows. Note that ``catalyst`` makes heavy usage of
rows. and print the first ten rows. Note that ``catalyst`` makes heavy usage of
`pandas <http://pandas.pydata.org/>`_, especially for data input and
outputting so it's worth spending some time to learn it.
@@ -265,17 +265,196 @@ outputting so it's worth spending some time to learn it.
perf = pd.read_pickle('buy_btc_simple_out.pickle') # read in perf DataFrame
perf.head()
.. raw:: html
<div style="max-height:1000px;max-width:1500px;overflow:auto;">
<table border="1" class="dataframe">
<thead>
<tr style="text-align: right;">
<th></th>
<th>algo_volatility</th>
<th>algorithm_period_return</th>
<th>alpha</th>
<th>benchmark_period_return</th>
<th>benchmark_volatility</th>
<th>beta</th>
<th>btc</th>
<th>capital_used</th>
<th>ending_cash</th>
<th>ending_exposure</th>
<th>...</th>
<th>short_exposure</th>
<th>short_value</th>
<th>shorts_count</th>
<th>sortino</th>
<th>starting_cash</th>
<th>starting_exposure</th>
<th>starting_value</th>
<th>trading_days</th>
<th>transactions</th>
<th>treasury_period_return</th>
</tr>
</thead>
<tbody>
<tr>
<th>2016-01-01 23:59:00+00:00</th>
<td>NaN</td>
<td>0.000000e+00</td>
<td>NaN</td>
<td>-0.010937</td>
<td>NaN</td>
<td>NaN</td>
<td>433.979999</td>
<td>0.000000</td>
<td>1.000000e+07</td>
<td>0.00</td>
<td>...</td>
<td>0</td>
<td>0</td>
<td>0</td>
<td>NaN</td>
<td>1.000000e+07</td>
<td>0.00</td>
<td>0.00</td>
<td>1</td>
<td>[]</td>
<td>0.0227</td>
</tr>
<tr>
<th>2016-01-02 23:59:00+00:00</th>
<td>0.000011</td>
<td>-9.536708e-07</td>
<td>-0.000170</td>
<td>-0.006480</td>
<td>0.173338</td>
<td>-0.000062</td>
<td>432.700000</td>
<td>-442.236708</td>
<td>9.999558e+06</td>
<td>432.70</td>
<td>...</td>
<td>0</td>
<td>0</td>
<td>0</td>
<td>-11.224972</td>
<td>1.000000e+07</td>
<td>0.00</td>
<td>0.00</td>
<td>2</td>
<td>[{u'order_id': u'7869f7828fa140328eb40477bb7de...</td>
<td>0.0227</td>
</tr>
<tr>
<th>2016-01-03 23:59:00+00:00</th>
<td>0.000011</td>
<td>-2.328842e-06</td>
<td>-0.000176</td>
<td>-0.026512</td>
<td>0.197857</td>
<td>0.000009</td>
<td>428.390000</td>
<td>-437.831716</td>
<td>9.999120e+06</td>
<td>856.78</td>
<td>...</td>
<td>0</td>
<td>0</td>
<td>0</td>
<td>-12.754262</td>
<td>9.999558e+06</td>
<td>432.70</td>
<td>432.70</td>
<td>3</td>
<td>[{u'order_id': u'be62ff77760c4599abaac43be9cc9...</td>
<td>0.0227</td>
</tr>
<tr>
<th>2016-01-04 23:59:00+00:00</th>
<td>0.000011</td>
<td>-2.380954e-06</td>
<td>-0.000139</td>
<td>-0.008640</td>
<td>0.269790</td>
<td>0.000020</td>
<td>432.900000</td>
<td>-442.441116</td>
<td>9.998677e+06</td>
<td>1298.70</td>
<td>...</td>
<td>0</td>
<td>0</td>
<td>0</td>
<td>-11.287205</td>
<td>9.999120e+06</td>
<td>856.78</td>
<td>856.78</td>
<td>4</td>
<td>[{u'order_id': u'd6dca79513214346a646079213526...</td>
<td>0.0224</td>
</tr>
<tr>
<th>2016-01-05 23:59:00+00:00</th>
<td>0.000011</td>
<td>-3.650729e-06</td>
<td>-0.000158</td>
<td>-0.021426</td>
<td>0.245989</td>
<td>0.000024</td>
<td>431.840000</td>
<td>-441.357754</td>
<td>9.998236e+06</td>
<td>1727.36</td>
<td>...</td>
<td>0</td>
<td>0</td>
<td>0</td>
<td>-12.333847</td>
<td>9.998677e+06</td>
<td>1298.70</td>
<td>1298.70</td>
<td>5</td>
<td>[{u'order_id': u'505275d6646a41f3856b22b16678d...</td>
<td>0.0225</td>
</tr>
</tbody>
</table>
</div>
|
There is a row for each trading day, starting on the first day of our
simulation Jan 1st, 2016. In the columns you can find various
information about the state of your algorithm. The very first column
information about the state of your algorithm. The column
``btc`` was placed there by the ``record()`` function mentioned earlier
and allows us to plot the price of bitcoin. For example, we could easily
examine now how our portfolio value changed over time compared to the
bitcoin price.
Our algorithm performance as assessed by the
``portfolio_value`` closely matches that of the bitcoin price. This
is not surprising as our algorithm only bought bitcoin every chance it got.
.. code-block:: python
%pylab inline
figsize(12, 12)
import matplotlib.pyplot as plt
ax1 = plt.subplot(211)
perf.portfolio_value.plot(ax=ax1)
ax1.set_ylabel('portfolio value')
ax2 = plt.subplot(212, sharex=ax1)
perf.btc.plot(ax=ax2)
ax2.set_ylabel('bitcoin price')
.. parsed-literal::
Populating the interactive namespace from numpy and matplotlib
.. parsed-literal::
<matplotlib.text.Text at 0x10eaeadd0>
.. image:: https://s3.amazonaws.com/enigmaco-docs/github.io/buy_btc_simple_graph.png
Our algorithm performance as assessed by the ``portfolio_value`` closely
matches that of the bitcoin price. This is not surprising as our algorithm
only bought bitcoin every chance it got.
Access to previous prices using ``history``
@@ -305,13 +484,14 @@ a function we use in the ``handle_data()`` section:
.. code-block:: python
from catalyst.api import order, record, symbol
%%catalyst --start 2016-1-1 --end 2017-9-30 -x bitfinex -o dma.pickle
from catalyst.api import order, record, symbol, order_target
def initialize(context):
context.i = 0
context.asset = symbol('btc_usd')
def handle_data(context, data):
def handle_data(context, data):
# Skip first 300 days to get full windows
context.i += 1
if context.i < 300:
@@ -336,6 +516,46 @@ a function we use in the ``handle_data()`` section:
short_mavg=short_mavg,
long_mavg=long_mavg)
def analyze(context, perf):
import matplotlib as plt
fig = plt.figure()
ax1 = fig.add_subplot(211)
perf.portfolio_value.plot(ax=ax1)
ax1.set_ylabel('portfolio value in $')
ax2 = fig.add_subplot(212)
perf['btc'].plot(ax=ax2)
perf[['short_mavg', 'long_mavg']].plot(ax=ax2)
perf_trans = perf.ix[[t != [] for t in perf.transactions]]
buys = perf_trans.ix[[t[0]['amount'] > 0 for t in perf_trans.transactions]]
sells = perf_trans.ix[
[t[0]['amount'] < 0 for t in perf_trans.transactions]]
ax2.plot(buys.index, perf.short_mavg.ix[buys.index],
'^', markersize=10, color='m')
ax2.plot(sells.index, perf.short_mavg.ix[sells.index],
'v', markersize=10, color='k')
ax2.set_ylabel('price in $')
plt.legend(loc=0)
plt.show()
Here we are explicitly defining an ``analyze()`` function that gets
automatically called once the backtest is done.
Although it might not be directly apparent, the power of ``history()``
(pun intended) can not be under-estimated as most algorithms make use of
prior market developments in one form or another. You could easily
devise a strategy that trains a classifier with
`scikit-learn <http://scikit-learn.org/stable/>`__ which tries to
predict future market movements based on past prices (note, that most of
the ``scikit-learn`` functions require ``numpy.ndarray``\ s rather than
``pandas.DataFrame``\ s, so you can simply pass the underlying
``ndarray`` of a ``DataFrame`` via ``.values``).
We also used the ``order_target()`` function above. This and other
functions like it can make order management and portfolio rebalancing
much easier.
Conclusions
~~~~~~~~~~~