Program PythonShell3_SVR_21;

//# -*- coding: utf-8 -*-  http://127.0.0.1:8080
//import scrapy - precondition note: change pyscript path of 991_oma_chartregression2.py

Const
 PYPATH='C:\Users\max\AppData\Local\Programs\Python\Python36-32\';
 PYPATH64='C:\Users\max\AppData\Local\Programs\Python\Python36\';
 
 PYCODE='C:\Users\max\SkyDrive\IBZ_Prozesstechnik_2016\hellomaxbox\.vscode\urlschemaload.py';
 PYSCRIPT= 'C:\maXbox\mX46210\DataScience\confusionlist\991_oma_chartregression2.py';
 
 PYFILE = 'input("prompt: ")'+CRLF+
          'def mygauss3():  '+CRLF+
          '#    i=0         '+CRLF+
          '     return sum(i for i in range(101))'+CRLF+
          '#    i=sum(i for i in range(101))'+CRLF+
          '#    print(i)   '+CRLF+ 
          '                '+CRLF+
          'print(mygauss3())'+CRLF+
          'k=input("press close to exit") '+CRLF+
          '#input("prompt: ")';
          
 PYFILE1    =//                "source": [
                'listOfNumbers = [1, 2, 3, 4, 5, 6]'+CRLF+
                                   ''+CRLF+
                                   'for number in listOfNumbers:'+CRLF+
                                   '    print(number)'+CRLF+
                                   '    if (number % 2 == 0):'+CRLF+
                                   '        print(''\"is even\"'')'+CRLF+
                                   '    else:'+CRLF+
                                   '        print(''\"is odd\"'')'+CRLF+
                                   '        '+CRLF+
                                   '    print (''\"All done.\"'')'+CRLF;

 PYFILE2 = ' def'+CRLF;

 PYFILE3 = 'sum(i for i in range(101))';
 
 PYFILE4 = 'def mygauss3():  '+CRLF+
          '#    i=0         '+CRLF+
          '     return sum(i for i in range(101))'+CRLF+
          '#    i=sum(i for i in range(101))'+CRLF+
          '#    print(i)   '+CRLF+ 
          '                '+CRLF+
          'print(mygauss3())'+CRLF+
          '';
          
PYSCRIPT5 = 
'import numpy as np'+CRLF+
'import matplotlib.pyplot as plt'+CRLF+
'import sys'+CRLF+
'from sklearn import tree'+CRLF+
'from sklearn.svm import SVC'+CRLF+
'from sklearn.ensemble import RandomForestClassifier'+CRLF+
'from sklearn.linear_model import LogisticRegression'+CRLF+
'from sklearn.preprocessing import StandardScaler'+CRLF+
'from sklearn.metrics import accuracy_score'+CRLF+
'from sklearn.model_selection import train_test_split'+CRLF+
' '+CRLF+
' '+CRLF+
'# Quotes from Yahoo finance and find optimal moving average'+CRLF+
'import pandas_datareader.data as web'+CRLF+
'  '+CRLF+
'#DataMax - Predict for 30 days; Predicted has data of Adj. Close shifted up by 30 rows'+CRLF+
'forecast_len=80  #default oma is 5'+CRLF+
'YQUOTES = "^SSMI"'+CRLF+
'PLOT = "Y"'+CRLF+
'try:'+CRLF+
'  forecast_len = int(sys.argv[1])'+CRLF+
'  #forecast_len= int(" ".join(sys.argv[1:]))'+CRLF+
'  YQUOTES = str(sys.argv[2])'+CRLF+
'  PLOT = str(sys.argv[3])'+CRLF+
'except:'+CRLF+
'  forecast_len= forecast_len'+CRLF+
'  YQUOTES = YQUOTES'+CRLF+
'  '+CRLF+  
'#YQUOTES = "BTC-USD"  #^GDAXI" , "^SSMI" , "^GSPC" (S&P 500 ) - ticker="GOOGL"'+CRLF+
'try: '+CRLF+
'  df= web.DataReader(YQUOTES, data_source="yahoo",start="09-11-2010")'+CRLF+  
'except:'+CRLF+
'  YQUOTES = "^SSMI"'+CRLF+
'  df= web.DataReader(YQUOTES, data_source="yahoo",start="09-11-2010")'+CRLF+
'  print("Invalid Quote Symbol got ^SSMI instead")'+CRLF+
'  '+CRLF+    
'#data = " ".join(sys.argv[1:])'+CRLF+
'print ("get forecast len:",forecast_len, "for ", YQUOTES)'+CRLF+
'quotes = df'+CRLF+
'print(quotes.info(5))'+CRLF+
'print(quotes["Close"][:5])'+CRLF+
'print(quotes["Close"][-3:])'+CRLF+
'df["_SMI_20"] = df.iloc[:,3].rolling(window=20).mean()'+CRLF+
'df["_SMI_60"] = df.iloc[:,3].rolling(window=60).mean()'+CRLF+
'df["_SMI_180"] = df.iloc[:,3].rolling(window=180).mean()'+CRLF+
'df["_SMI_OMA"] = df.iloc[:,3].rolling(window=forecast_len).mean()'+CRLF+
'  '+CRLF+
'#"""'+CRLF+
'if PLOT=="Y":'+CRLF+
'   x_ax_time = quotes.index #range(len(df))'+CRLF+
'   plt.figure(figsize=[12,7])'+CRLF+
'   plt.grid(True)'+CRLF+
'   plt.title("Optimal Moving Average OMA: "+YQUOTES, fontsize=18)'+CRLF+
'   plt.plot(x_ax_time, quotes["Close"], label="Close")'+CRLF+
'   plt.plot(x_ax_time, df["_SMI_60"],label="MA 3 Month")'+CRLF+
'   plt.plot(x_ax_time, df["_SMI_180"],label="MA 9 Month")'+CRLF+
'   plt.plot(x_ax_time, df["_SMI_OMA"],label= "OMA "+str(forecast_len)+" D.")'+CRLF+
'   #plt.xlabel("days", fontsize=15)'+CRLF+
'   # plt.plot_date(quotes.index, quotes["Close"])'+CRLF+
'   plt.legend(loc=2)'+CRLF+
'   plt.show()'+CRLF+
'#"""'+CRLF+
'  '+CRLF+
'dates = quotes.index'+CRLF+
'dates = dates[1:]'+CRLF+
'#closing_values = np.array([quote[3] for quote in quotes])'+CRLF+
'#volume_of_shares = np.array([quote[5] for quote in quotes])[1:]'+CRLF+
'closing_values = np.array(quotes["Close"])'+CRLF+
'volume_of_shares = np.array(quotes["Volume"])'+CRLF+
'  '+CRLF+
'#Predict for 30 days; Predicted has Quotes of Close shifted up by 30 rows'+CRLF+
'ytarget= quotes["Close"].shift(-forecast_len)'+CRLF+
'ytarget= ytarget[:-forecast_len]'+CRLF+
'Xdata= closing_values[:-forecast_len]'+CRLF+
'#print("Offset shift:",ytarget[:10])'+CRLF+
'  '+CRLF+
'# Feature Scaling'+CRLF+
'#sc_X = StandardScaler()'+CRLF+
'#sc_y = StandardScaler()'+CRLF+
'#Xdata = sc_X.fit_transform(Xdata.reshape(-1,1))'+CRLF+
'#You need to do this is that pandas Series objects are by design one dimensional.'+CRLF+
'#ytarget = sc_y.fit_transform(ytarget.values.reshape(-1,1))'+CRLF+
'   '+CRLF+
'from sklearn.svm import SVR'+CRLF+
'# Split datasets into training and test sets (80% and 20%)'+CRLF+
'print("target shape len2: ",len(ytarget),len(Xdata))'+CRLF+
'x_train,x_test,y_train,y_test=train_test_split(Xdata,ytarget,test_size=0.2, \'+CRLF+
'                                                       random_state= 72)'+CRLF+
'print("xtrain shape len3: ",len(x_train),len(y_train))'+CRLF+
'   '+CRLF+
'# - Create SVR model and train it'+CRLF+
'svr_rbf= SVR(kernel="rbf",C=1e3,gamma=0.1)'+CRLF+ 
'x_train = x_train.reshape(-1,1)'+CRLF+
'svr_rbf.fit(x_train,y_train)'+CRLF+
'   '+CRLF+
'# Predicting single value as new result'+CRLF+
'print("predict old in :", forecast_len, svr_rbf.predict([quotes["Close"][:1]]))'+CRLF+
'print("predict now in :", forecast_len, svr_rbf.predict([quotes["Close"][-1:]]))'+CRLF+
'    '+CRLF+
'#DBASTAr - Get score'+CRLF+
'svr_rbf_confidence=svr_rbf.score(x_test.reshape(-1,1),y_test)'+CRLF+
'print(f"SVR Confidence: {round(svr_rbf_confidence*100,2)}%")';
          
 ACTIVESCRIPT = PYSCRIPT5;
 
 function CompareFilename(List: TStringList; Index1, Index2: Integer): Integer;
var
  fn1, fn2: String;
begin
  fn1 := List.Names[Index1];
  fn2 := List.Names[Index2];
  Result := CompareString(fn1, fn2);
end;

function CompareFileSize(List: TStringList; Index1, Index2: Integer): Integer;
var
  sz1, sz2: Int64;
begin
  sz1 := StrToInt(List.ValueFromIndex[Index1]);
  sz2 := StrToInt(List.ValueFromIndex[Index2]);
  Result := ord(CompareValueI(sz1, sz2));
end;

Function GetValueFromIndex(R: TStringList; Index: Integer):String;
var
  S: string;
  i: Integer;
begin
  S := R.Strings[Index];
  i := Pos('=', S);
  if I > 0 then
    result := Copy(S, i+1, MaxInt)
  else
    result := '';
end;

Function dummy(Reqlist: TStringList):String;
var
  i: Integer;
  RESULTv: string;
begin
  for i := 0 to ReqList.Count-1 do
    RESULTv := RESULTv + Reqlist.Names[i] + ' -> ' + GetValueFromIndex(Reqlist, i);
  result := RESULTv;
end;

 var fcast: integer;
     olist: TStringlist;
     dosout, theMaxOMA, QuoteSymbol: string;
     Yvalfloat: array[1..500] of double; //TDynfloatArray;
     //theMaxFloat: double;
     
     RUNSCRIPT: string;

begin //@main

  //saveString(exepath+'mygauss.py',ACTIVESCRIPT);
  saveString(exepath+'991_oma_chartregression5.py',ACTIVESCRIPT);
  sleep(300)
  //if fileExists(PYPATH+'python.exe') then 
   //if fileExists(PYSCRIPT) then begin
   if fileExists(exepath+'991_oma_chartregression5.py') then begin
      RUNSCRIPT:= exepath+'991_oma_chartregression5.py';
     //ShellExecute3('cmd','/k '+PYPATH+'python.exe && '+PYFILE +'&& mygauss3()'
     //                                  ,secmdopen);
      { ShellExecute3('cmd','/k '+PYPATH+
                        'python.exe && exec(open('+exepath+'mygauss.py'').read())'
                        ,secmdopen);
                 }       
     { ShellExecute3('cmd','/k '+PYPATH+
                        'python.exe '+exepath+'mygauss.py', secmdopen); }
     // ShellExecute3(PYPATH+'python.exe ',exepath+'mygauss.py'
     //                   ,secmdopen);
     maxform1.console1click(self);
     memo2.height:= 205;
     // maxform1.shellstyle1click(self);
     // writeln(GetDosOutput(PYPATH+'python.exe '+PYSCRIPT,'C:\'));
    fcast:= 120;      //first forecast with plot
    QuoteSymbol:= 'BTC-USD'; //'BTC-USD'; //^SSMI
    olist:= TStringlist.create;
    olist.NameValueSeparator:= '=';
    //olist.Sorted:= True;
    //olist.CustomSort(@CompareFileName)
    //GetDosOutput('py '+PYSCRIPT+' '+itoa(fcast)+' '+QuoteSymbol+' "Y"','C:\');
    GetDosOutput('py '+RUNSCRIPT+' '+itoa(fcast)+' '+QuoteSymbol+' "Y"','C:\');
  
    for it:= 20 to 130 do 
       if it mod 5=0 then begin
         //(GetDosOutput('py '+PYSCRIPT+' '+itoa(it)+' "BTC-USD"'+ 'Plot?','C:\'));
         dosout:= GetDosOutput('py '+RUNSCRIPT+' '+itoa(it)+' '+QuoteSymbol+' "N"','C:\');  
         writeln(dosout)
         with TRegExpr.Create do begin
            //Expression:=('SVR Confidence: ([0-9\.%]+).*');
            Expression:=('SVR Confidence: ([0-9\.]+).*');
            if Exec(dosout) then begin
              PrintF('Serie %d  : %s',[it, Match[1]]);
              olist.add(Match[1]+'='+itoa(it));
              Yvalfloat[it]:= strtofloat(Copy(match[1],1,5));  
              //MaxFloatArray           
            end;
          Free;
         end; 
       end;   
     writeln(CR+LF+olist.text)
     writeln('OMA from key value list2: '+floattostr(MaxFloatArray(Yvalfloat)))
     TheMaxOMA:= olist.Values[floattostr(MaxFloatArray(Yvalfloat))];
     writeln('OMA for Chart Signal: '+TheMaxOMA);
     olist.Free;  
    (GetDosOutput('py '+RUNSCRIPT+' '+(TheMaxOMA)+' '+QuoteSymbol+' "Y"','C:\'));
   end;
End.

ref: https://docs.scrapy.org/en/latest/topics/link-extractors.html

https://doc.scrapy.org/en/latest/topics/architecture.html

https://www.rosettacode.org/wiki/Cumulative_standard_deviation#Pascal

https://www.kaggle.com/vsmolyakov/keras-cnn-with-fasttext-embeddings

https://www.angio.net/pi/bigpi.cgi

https://medium.com/analytics-vidhya/using-deep-learning-to-create-a-stock-trading-bot-a96e6351d31c

This function changes all data into a value between 0 and 1. This is as many stocks have skyrocketed or nosedived. Without normalizing, the neural network would learn from datapoints with higher values. This could create a blind spot and therefore affect predictions. The normalizing is done as so:

value = (value - minimum) / maximum

if Exec(memo2.text) then
         for i:=0 to SubExprMatchCount do
            PrintF('Group %d  : %s', [i, Match[i]]);


class ExampleSpider(scrapy.Spider):
    name = 'example'
    #allowed_domains = ['www.ibm.ch']
    allowed_domains = ['127.0.0.1']
  
    #start_urls = ['http://www.ibm.ch/']
    start_urls = ['http://127.0.0.1:8080']
                                       
    def parse(self, response):
        pass        
        
        
        
        "source": [
    "import numpy as np\n",
    "\n",
    "A = np.random.normal(25.0, 5.0, 10)\n",
    "print (A)"
   ]
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "## Activity"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "deletable": true,
    "editable": true
   },
   "source": [
    "Write some code that creates a list of integers, loops through each element of the list, and only prints out even numbers!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "deletable": true,
    "editable": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}