Merge branch 'matplotlib:master' into master

Author: DanielGoldfarb

examples/panels.ipynb

@@ -17,7 +17,7 @@
     "![sample_subplot](data/sample_subplot.png)\n",
     "\n",
     "### Features and Basic Terminology\n",
-    "* Panels are identified by their \"Panel ID\", an integer ranging from 0 to 9.\n",
+    "* Panels are identified by their \"Panel ID\", an integer ranging from 0 to 31.\n",
     "* Panel ID's are always numbered from top to bottom, thus:\n",
     "  - Panel 0 is always the uppermost panel, Panel 1 is just below that, and so on.\n",
     "* The \"*main panel*\" is the panel where candlesticks/ohlc data are plotted.\n",

markdown/subplots.md

@@ -17,10 +17,10 @@
 * The Panels Method attains its simplicity, in part, by having certain limitations.<br>These limitiations are:
    - Subplots are always stacked vertically.
    - All subplots share the same x-axis.
-   - There is a maximum of 10 subplots.
+   - There is a maximum of 32 subplots.
 * The Panels Method is adequate to plot:
   - ohlc, candlesticks, etc.
-  - with volume, and 
+  - with volume, and
   - with one or more studies/indicators, such as:
     - MACD, DMI, RSI, Bollinger, Accumulation/Distribution Oscillator, Commodity Channel Index, Etc.
 * [**See here for a tutorial and details on implementing the mplfinance Panels Method for subplots.**](https://github.com/matplotlib/mplfinance/blob/master/examples/panels.ipynb)

src/mplfinance/_utils.py

@@ -1,4 +1,4 @@
-version_info = (0, 12, 9, 'beta', 1)
+version_info = (0, 12, 9, 'beta', 3)
 
 _specifier_ = {'alpha': 'a','beta': 'b','candidate': 'rc','final': ''}
 

src/mplfinance/_version.py

@@ -119,7 +119,17 @@ def _valid_plot_kwargs():
         'mav'                       : { 'Default'     : None,
                                         'Description' : 'Moving Average window size(s); (int or tuple of ints)',
                                         'Validator'   : _mav_validator },
+
+        'ema'                       : { 'Default'     : None,
+                                        'Description' : 'Exponential Moving Average window size(s); (int or tuple of ints)',
+                                        'Validator'   : _mav_validator },
 
+        'mavcolors'                 : { 'Default'     : None,
+                                        'Description' : 'color cycle for moving averages (list or tuple of colors)'+
+                                                        '(overrides mpf style mavcolors).',
+                                        'Validator'   : lambda value: isinstance(value,(list,tuple)) and
+                                                                      all([mcolors.is_color_like(v) for v in value]) },
+
         'renko_params'              : { 'Default'     : dict(),
                                         'Description' : 'dict of renko parameters; call `mpf.kwarg_help("renko_params")`',
                                         'Validator'   : lambda value: isinstance(value,dict) },
@@ -450,6 +460,13 @@ def plot( data, **kwargs ):
     else:
         raise TypeError('style should be a `dict`; why is it not?')
 
+    if config['mavcolors'] is not None:
+        config['_ma_color_cycle'] = cycle(config['mavcolors'])
+    elif style['mavcolors'] is not None:
+        config['_ma_color_cycle'] = cycle(style['mavcolors'])
+    else:
+        config['_ma_color_cycle'] = None
+
     if not external_axes_mode:
         fig = plt.figure()
         _adjust_figsize(fig,config)
@@ -528,8 +545,10 @@ def plot( data, **kwargs ):
 
     if ptype in VALID_PMOVE_TYPES:
         mavprices = _plot_mav(axA1,config,xdates,pmove_avgvals)
+        emaprices = _plot_ema(axA1, config, xdates, pmove_avgvals)
     else:
         mavprices = _plot_mav(axA1,config,xdates,closes)
+        emaprices = _plot_ema(axA1, config, xdates, closes)
 
     avg_dist_between_points = (xdates[-1] - xdates[0]) / float(len(xdates))
     if not config['tight_layout']:
@@ -595,6 +614,13 @@ def plot( data, **kwargs ):
             else:
                 for jj in range(0,len(mav)):     
                     retdict['mav' + str(mav[jj])] = mavprices[jj]
+        if config['ema'] is not None:
+            ema = config['ema']
+            if len(ema) != len(emaprices):
+                warnings.warn('len(ema)='+str(len(ema))+' BUT len(emaprices)='+str(len(emaprices)))
+            else:
+                for jj in range(0, len(ema)):
+                    retdict['ema' + str(ema[jj])] = emaprices[jj]
         retdict['minx'] = minx
         retdict['maxx'] = maxx
         retdict['miny'] = miny
@@ -1129,10 +1155,7 @@ def _plot_mav(ax,config,xdates,prices,apmav=None,apwidth=None):
         if len(mavgs) > 7:
             mavgs = mavgs[0:7]  # take at most 7
 
-        if style['mavcolors'] is not None:
-            mavc = cycle(style['mavcolors'])
-        else:
-            mavc = None
+        mavc = config['_ma_color_cycle']
 
         for idx,mav in enumerate(mavgs):
             mean = pd.Series(prices).rolling(mav).mean()
@@ -1147,6 +1170,42 @@ def _plot_mav(ax,config,xdates,prices,apmav=None,apwidth=None):
             mavp_list.append(mavprices)
     return mavp_list
 
+
+def _plot_ema(ax,config,xdates,prices,apmav=None,apwidth=None):
+    '''ema: exponential moving average'''
+    style = config['style']
+    if apmav is not None:
+        mavgs = apmav
+    else:
+        mavgs = config['ema']
+    mavp_list = []
+    if mavgs is not None:
+        shift = None
+        if isinstance(mavgs,dict):
+            shift = mavgs['shift']
+            mavgs = mavgs['period']
+        if isinstance(mavgs,int):
+            mavgs = mavgs,      # convert to tuple
+        if len(mavgs) > 7:
+            mavgs = mavgs[0:7]  # take at most 7
+
+        mavc = config['_ma_color_cycle']
+
+        for idx,mav in enumerate(mavgs):
+            # mean = pd.Series(prices).rolling(mav).mean()
+            mean = pd.Series(prices).ewm(span=mav,adjust=False).mean()
+            if shift is not None:
+                mean = mean.shift(periods=shift[idx])
+            emaprices = mean.values
+            lw = config['_width_config']['line_width']
+            if mavc:
+                ax.plot(xdates, emaprices, linewidth=lw, color=next(mavc))
+            else:
+                ax.plot(xdates, emaprices, linewidth=lw)
+            mavp_list.append(emaprices)
+    return mavp_list
+
+
 def _auto_secondary_y( panels, panid, ylo, yhi ):
     # If mag(nitude) for this panel is not yet set, then set it
     # here, as this is the first ydata to be plotted on this panel:

src/mplfinance/plotting.py

@@ -0,0 +1,124 @@
+import os
+import os.path
+import glob
+import mplfinance as mpf
+import pandas as pd   
+import matplotlib.pyplot as plt
+from   matplotlib.testing.compare import compare_images
+
+print('mpf.__version__ =',mpf.__version__)                 # for the record
+print('mpf.__file__ =',mpf.__file__)                       # for the record
+print("plt.rcParams['backend'] =",plt.rcParams['backend']) # for the record
+
+base='ema'
+tdir = os.path.join('tests','test_images')
+refd = os.path.join('tests','reference_images')
+
+globpattern  = os.path.join(tdir,base+'*.png')
+oldtestfiles = glob.glob(globpattern)
+for fn in oldtestfiles:
+    try:
+        os.remove(fn)
+    except:
+        print('Error removing file "'+fn+'"')
+
+IMGCOMP_TOLERANCE = 10.0  # this works fine for linux
+# IMGCOMP_TOLERANCE = 11.0  # required for a windows pass. (really 10.25 may do it).
+
+_df = pd.DataFrame()
+def get_ema_data():
+    global _df
+    if len(_df) == 0:
+        _df = pd.read_csv('./examples/data/yahoofinance-GOOG-20040819-20180120.csv',
+                          index_col='Date',parse_dates=True)
+    return _df
+    
+
+def create_ema_image(tname):
+
+    df = get_ema_data()
+    df = df[-50:]               # show last 50 data points only                     
+
+    ema25 = df['Close'].ewm(span=25.0, adjust=False).mean()       
+    mav25 = df['Close'].rolling(window=25).mean()
+
+    ap = [
+        mpf.make_addplot(df, panel=1, type='ohlc', color='c',
+                         ylabel='mpf mav', mav=25, secondary_y=False),
+        mpf.make_addplot(ema25, panel=2, type='line', width=2,  color='c',
+                         ylabel='calculated', secondary_y=False),
+        mpf.make_addplot(mav25, panel=2, type='line', width=2, color='blue',
+                         ylabel='calculated', secondary_y=False)
+    ]
+
+    # plot and save in `tname` path
+    mpf.plot(df, ylabel="mpf ema", type='ohlc',
+             ema=25, addplot=ap, panel_ratios=(1, 1), savefig=tname
+    )
+
+
+def test_ema01():
+
+    fname = base+'01.png'
+    tname = os.path.join(tdir,fname)
+    rname = os.path.join(refd,fname)
+
+    create_ema_image(tname)
+
+    tsize = os.path.getsize(tname)
+    print(glob.glob(tname),'[',tsize,'bytes',']')
+
+    rsize = os.path.getsize(rname)
+    print(glob.glob(rname),'[',rsize,'bytes',']')
+
+    result = compare_images(rname,tname,tol=IMGCOMP_TOLERANCE)
+    if result is not None:
+        print('result=',result)
+    assert result is None
+
+def test_ema02():
+    fname = base+'02.png'
+    tname = os.path.join(tdir,fname)
+    rname = os.path.join(refd,fname)
+
+    df = get_ema_data()
+    df = df[-125:-35]
+
+    mpf.plot(df, type='candle', ema=(5,15,25), mav=(5,15,25), savefig=tname)
+
+    tsize = os.path.getsize(tname)
+    print(glob.glob(tname),'[',tsize,'bytes',']')
+
+    rsize = os.path.getsize(rname)
+    print(glob.glob(rname),'[',rsize,'bytes',']')
+
+    result = compare_images(rname,tname,tol=IMGCOMP_TOLERANCE)
+    if result is not None:
+        print('result=',result)
+    assert result is None
+
+def test_ema03():
+    fname = base+'03.png'
+    tname = os.path.join(tdir,fname)
+    rname = os.path.join(refd,fname)
+
+    df = get_ema_data()
+    df = df[-125:-35]
+
+    mac = ['red','orange','yellow','green','blue','purple']
+
+    mpf.plot(df, type='candle', ema=(5,10,15,25), mav=(5,15,25),
+             mavcolors=mac, savefig=tname)
+
+
+    tsize = os.path.getsize(tname)
+    print(glob.glob(tname),'[',tsize,'bytes',']')
+
+    rsize = os.path.getsize(rname)
+    print(glob.glob(rname),'[',rsize,'bytes',']')
+
+    result = compare_images(rname,tname,tol=IMGCOMP_TOLERANCE)
+    if result is not None:
+        print('result=',result)
+    assert result is None
+