save a few more experiments/examples

Author: DanielGoldfarb

examples/scratch_pad/lines.dill

@@ -0,0 +1,98 @@
+import pandas as pd
+import mplfinance as mpf
+import dill
+import os
+from matplotlib.widgets import MultiCursor
+
+# read the data:
+idf = pd.read_csv('../data/SPY_20110701_20120630_Bollinger.csv',index_col=0,parse_dates=True)
+df  = idf.loc['2011-07-01':'2011-12-30',:]
+
+
+# macd related calculations:
+exp12 = df['Close'].ewm(span=12, adjust=False).mean()
+exp26 = df['Close'].ewm(span=26, adjust=False).mean()
+macd = exp12 - exp26
+signal    = macd.ewm(span=9, adjust=False).mean()
+histogram = macd - signal
+
+# initial plot:
+apds = [mpf.make_addplot(exp12,color='lime'),
+        mpf.make_addplot(exp26,color='c'),
+        mpf.make_addplot(histogram,type='bar',width=0.7,panel=1,
+                         color='dimgray',alpha=1,secondary_y=False),
+        mpf.make_addplot(macd,panel=1,color='fuchsia',secondary_y=True),
+        mpf.make_addplot(signal,panel=1,color='b',secondary_y=True),
+       ]
+
+# For some reason, which i have yet to determine, MultiCursor somehow
+# causes ymin to be set to zero for the main candlestick Axes, but we
+# can correct that problem by passing in specific values:
+ymin = min(df['Low'])  * 0.98
+ymax = max(df['High']) * 1.02
+
+# initial plot with cursor:
+if os.path.exists('lines.dill'):
+    alines = dill.load(open('lines.dill','rb'))
+    fig, axlist = mpf.plot(df,type='candle',addplot=apds,figscale=1.25,figratio=(8,6),title='\nMACD', ylim=(ymin,ymax),
+                           alines=dict(alines=alines,colors='r'),
+                           style='blueskies',volume=True,volume_panel=2,panel_ratios=(6,3,2),returnfig=True)
+else:
+    alines = []
+    fig, axlist = mpf.plot(df,type='candle',addplot=apds,figscale=1.25,figratio=(8,6),title='\nMACD', ylim=(ymin,ymax),
+                           alines=dict(alines=alines,colors='r'),
+                           style='blueskies',volume=True,volume_panel=2,panel_ratios=(6,3,2),returnfig=True)
+multi = MultiCursor(fig.canvas, axlist[0:2], horizOn=True, vertOn=True, color='pink', lw=1.2)
+
+fig.canvas.draw_idle()
+
+# ---------------------------------------------------
+# set up an event loop where we wait for two
+# mouse clicks, and then draw a line in between them,
+# and then wait again for another two mouse clicks.
+
+# This is a crude way to do it, but its quick and easy.
+# Disadvantage is: user has 8 seconds to provide two clicks
+# or the first click will be erased.  But the 8 seconds
+# repeats as long as the user does not close the Figure,
+# so user can draw as many trend lines as they want.
+# The advantage of doing it this way is we don't have
+# to write all the mouse click handling stuff that's
+# already written in `Figure.ginput()`.
+
+
+not_closed = True
+def on_close(event):
+    global not_closed
+    global alines
+    dill.dump(alines, open('lines.dill','wb'))   
+    not_closed = False
+
+fig.canvas.mpl_connect('close_event', on_close)
+
+while not_closed:
+
+    vertices = fig.ginput(n=2,timeout=8)
+    if len(vertices) < 2:
+        continue
+    p1 = vertices[0]
+    p2 = vertices[1]
+
+    d1 = df.index[ round(p1[0]) ]
+    d2 = df.index[ round(p2[0]) ]
+
+    alines.append( [ (d1,p1[1]), (d2,p2[1]) ] )
+
+    apds = [mpf.make_addplot(exp12,color='lime',ax=axlist[0]),
+            mpf.make_addplot(exp26,color='c',ax=axlist[0]),
+            mpf.make_addplot(histogram,type='bar',width=0.7,panel=1,ax=axlist[2],color='dimgray',alpha=1),
+            mpf.make_addplot(macd,panel=1,color='fuchsia',ax=axlist[3]),
+            mpf.make_addplot(signal,panel=1,color='b',ax=axlist[3])
+           ]
+
+    mpf.plot(df,ax=axlist[0],type='candle',addplot=apds,ylim=(ymin,ymax),
+             alines=dict(alines=alines,colors='r'),
+             style='blueskies',volume=axlist[4],volume_panel=2,panel_ratios=(6,3,2))
+
+    fig.canvas.draw_idle()
+