01_aiif_webinar.md

AI in Finance: An Introduction in Python

Dr Yves J Hilpisch | The Python Quants & The AI Machine

Webinar, DataCamp, 21. May 2019

Short Link

http://bit.ly/aiif_webinar

Slides

http://hilpisch.com/aiif_webinar.pdf

Python Certificate Programs

• http://certificate.tpq.io (University Certificate in Python for Algorithmic Trading)
• http://compfinance.tpq.io (University Certificate in Python for Computational Finance)
• http://platinum.tpq.io (Combination of the above two University Certificates)

Python for Finance (2nd ed.)

Learn all about my most recent book under http://py4fi.tpq.io.

Sign up under http://py4fi.pqp.io to access and execute all Jupyter Notebooks and codes on our Quant Platform.

General Resources

• http://tpq.io (company page)
• http://hilpisch.com (personal page)
• http://twitter.com/dyjh (Twitter feed)

02_market_prediction.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src='http://hilpisch.com/taim_logo.png' width=\"350px\" align=\"right\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# AI in Finance\n",
    "\n",
    "**Market Prediction**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "&copy; Dr Yves J Hilpisch | The Python Quants GmbH\n",
    "\n",
    "http://tpq.io | http://twitter.com/dyjh"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from pylab import plt\n",
    "plt.style.use('seaborn')\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "url = 'http://hilpisch.com/oanda_eur_usd.csv'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "raw = pd.read_csv(url, index_col=0,\n",
    "                  parse_dates=True).dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "DatetimeIndex: 15456 entries, 2018-01-01 22:00:00 to 2019-03-29 20:30:00\n",
      "Data columns (total 6 columns):\n",
      "c           15456 non-null float64\n",
      "complete    15456 non-null bool\n",
      "h           15456 non-null float64\n",
      "l           15456 non-null float64\n",
      "o           15456 non-null float64\n",
      "volume      15456 non-null int64\n",
      "dtypes: bool(1), float64(4), int64(1)\n",
      "memory usage: 739.6 KB\n"
     ]
    }
   ],
   "source": [
    "raw.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The Features"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following creates a **set of financial features**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = raw.copy()\n",
    "\n",
    "# log returns & direction\n",
    "data['r'] = np.log(data['c'] / data['c'].shift(1))\n",
    "data['rs'] = (data['r'] - data['r'].mean()) / data['r'].std()\n",
    "data['d'] = np.where(data['r'] > 0, 1, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# rolling statistics\n",
    "data['v1'] = data['r'].rolling(20).std()\n",
    "data['v2'] = data['r'].rolling(100).std()\n",
    "data['sma1'] = data['c'].rolling(20).mean()\n",
    "data['sma2'] = data['c'].rolling(100).mean()\n",
    "data['mom1'] = data['r'].rolling(5).mean()\n",
    "data['mom2'] = data['r'].rolling(20).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "data.dropna(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "features = list(data.columns)\n",
    "features.remove('complete')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "15356"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ld = len(data)\n",
    "ld"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The data set is split into **train, validation and test data sets**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "split = int(len(data) * 0.7)\n",
    "val_size = int(split * 0.15)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "train = data.iloc[:split]\n",
    "val = train[-val_size:]\n",
    "train = train[:-val_size]\n",
    "test = data.iloc[split:].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "lags = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "def gaussian(x):\n",
    "    mean = x.mean()\n",
    "    std = x.std()\n",
    "    return (x - mean) / std, mean, std"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The following function creates **lags** of the features columns and **normalizes** the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def normalize_and_lag():\n",
    "    global cols\n",
    "    cols = []\n",
    "    for f in features:\n",
    "        for lag in range(1, lags + 1):\n",
    "            col = f'{f}_lag_{lag}'\n",
    "            if f in ['r', 'rs', 'd', 'u-d']:\n",
    "                train[col] = train[f].shift(lag)\n",
    "                val[col] = val[f].shift(lag)\n",
    "                test[col] = test[f].shift(lag)\n",
    "            else:\n",
    "                train[col], mean, std = gaussian(train[f].shift(lag))\n",
    "                val[col] = (val[f].shift(lag) - mean) / std\n",
    "                test[col] = (test[f].shift(lag) - mean) / std\n",
    "            cols.append(col)\n",
    "    train.dropna(inplace=True)\n",
    "    val.dropna(inplace=True)\n",
    "    test.dropna(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "normalize_and_lag()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "140"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cols)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# train.head(5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Backtesting"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Scikit-Learn"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, a `MLPClassifier` is trained."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.neural_network import MLPClassifier"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = MLPClassifier(hidden_layer_sizes=(256, 256),\n",
    "                      activation='relu',\n",
    "                      alpha=0.0001,\n",
    "                      random_state=100,\n",
    "                      max_iter=200,\n",
    "                      validation_fraction=0.0,\n",
    "                      shuffle=False,\n",
    "                      early_stopping=False,\n",
    "                      verbose=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 41s, sys: 9.83 s, total: 1min 51s\n",
      "Wall time: 18.6 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "MLPClassifier(activation='relu', alpha=0.0001, batch_size='auto', beta_1=0.9,\n",
       "       beta_2=0.999, early_stopping=False, epsilon=1e-08,\n",
       "       hidden_layer_sizes=(256, 256), learning_rate='constant',\n",
       "       learning_rate_init=0.001, max_iter=200, momentum=0.9,\n",
       "       n_iter_no_change=10, nesterovs_momentum=True, power_t=0.5,\n",
       "       random_state=100, shuffle=False, solver='adam', tol=0.0001,\n",
       "       validation_fraction=0.0, verbose=False, warm_start=False)"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time model.fit(train[cols], train['d'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Second, the trained model is used to predict the **future market direction**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "test['p'] = model.predict(test[cols])\n",
    "test['p'] = np.where(test['p'] > 0, 1, -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "test['s'] = test['p'] * test['r']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A look at the **positions** and the **number of trades**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " 1    3008\n",
       "-1    1589\n",
       "Name: p, dtype: int64"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test['p'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1764"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(test['p'].diff() != 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model **outperforms** the passive benchmark investment by a few percentage points.\n",
    "\n",
    "<b style=\"color:red;\">CAUTION: The analysis is implemented under the assumption of zero transaction costs.<b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "r    0.992410\n",
       "s    1.046602\n",
       "dtype: float64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test[['r', 's']].sum().apply(np.exp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "r    0.000602\n",
       "s    0.000602\n",
       "dtype: float64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test[['r', 's']].std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "test[['r', 's']].cumsum().apply(np.exp).plot(figsize=(10, 6));"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Keras"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A dense neural network with `Keras` is built and trained."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "tf.logging.set_verbosity(tf.logging.ERROR)\n",
    "from keras.layers import Dense, Dropout\n",
    "from keras.models import Sequential\n",
    "from keras.regularizers import l2\n",
    "from keras.callbacks import EarlyStopping"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(100)\n",
    "tf.random.set_random_seed(100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "\n",
    "model.add(Dense(128, activation='relu',\n",
    "                kernel_regularizer=l2(0.001),\n",
    "                input_shape=(len(cols),)\n",
    "               )\n",
    "         )\n",
    "model.add(Dropout(0.3, seed=100))\n",
    "model.add(Dense(128, activation='relu',\n",
    "                kernel_regularizer=l2(0.001)\n",
    "               )\n",
    "         )\n",
    "model.add(Dropout(0.3, seed=100))\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "model.compile(optimizer='adam',\n",
    "              loss='binary_crossentropy',\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense_1 (Dense)              (None, 128)               18048     \n",
      "_________________________________________________________________\n",
      "dropout_1 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 128)               16512     \n",
      "_________________________________________________________________\n",
      "dropout_2 (Dropout)          (None, 128)               0         \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 1)                 129       \n",
      "=================================================================\n",
      "Total params: 34,689\n",
      "Trainable params: 34,689\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "callbacks = [EarlyStopping(monitor='val_acc', patience=25)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 21.5 s, sys: 6.77 s, total: 28.2 s\n",
      "Wall time: 10.6 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1a39cac828>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%%time\n",
    "model.fit(train[cols], train['d'],\n",
    "          epochs=250, batch_size=32, verbose=False,\n",
    "          validation_data=(val[cols], val['d']),\n",
    "          callbacks=callbacks);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A look at how the **metrics** evolve over the training periods."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "res = pd.DataFrame(model.history.history)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>val_loss</th>\n",
       "      <th>val_acc</th>\n",
       "      <th>loss</th>\n",
       "      <th>acc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>0.693369</td>\n",
       "      <td>0.530587</td>\n",
       "      <td>0.692790</td>\n",
       "      <td>0.522187</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0.695514</td>\n",
       "      <td>0.503121</td>\n",
       "      <td>0.693074</td>\n",
       "      <td>0.521749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>0.697411</td>\n",
       "      <td>0.503745</td>\n",
       "      <td>0.692624</td>\n",
       "      <td>0.526679</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    val_loss   val_acc      loss       acc\n",
       "29  0.693369  0.530587  0.692790  0.522187\n",
       "30  0.695514  0.503121  0.693074  0.521749\n",
       "31  0.697411  0.503745  0.692624  0.526679"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "res.tail(3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "res.plot(figsize=(10, 6), style=['--', '--', '-', '-']);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The trained model is used to generate **directional market predictions** on the test data set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4597/4597 [==============================] - 0s 13us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.6929880122218361, 0.522732216779735]"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(test[cols], test['d'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "test['p'] = model.predict_classes(test[cols])\n",
    "test['p'] = np.where(test['p'] > 0, 1, -1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "test['s'] = test['p'] * test['r']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A look at the **positions** and the **number of trades**."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       " 1    3042\n",
       "-1    1555\n",
       "Name: p, dtype: int64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test['p'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1123"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sum(test['p'].diff() != 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model **outperforms** the passive benchmark investment.\n",
    "\n",
    "<b style=\"color:red;\">CAUTION: The analysis is implemented under the assumption of zero transaction costs.<b>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "r    0.992410\n",
       "s    1.078215\n",
       "dtype: float64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test[['r', 's']].sum().apply(np.exp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "r    0.000602\n",
       "s    0.000602\n",
       "dtype: float64"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test[['r', 's']].std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "test[['r', 's']].cumsum().apply(np.exp).plot(figsize=(10, 6));"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"http://hilpisch.com/tpq_logo.png\" alt=\"The Python Quants\" width=\"35%\" align=\"right\" border=\"0\"><br>\n",
    "\n",
    "<a href=\"http://tpq.io\" target=\"_blank\">http://tpq.io</a> | <a href=\"http://twitter.com/dyjh\" target=\"_blank\">@dyjh</a> | <a href=\"mailto:training@tpq.io\">training@tpq.io</a>"
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}

I got Tensorflow 2. Anyone has ideas on how to fix to work with TF 2? Dont want to downgrade to TF 1.14. Thanks.

I got Tensorflow 2. Anyone has ideas on how to fix to work with TF 2? Dont want to downgrade to TF 1.14. Thanks.

What issues do you face? Tensorflow is used only indirectly via Keras.

When executing some lines of code, Python recommends to use tf.keras or to use tensorflow 1.14.

…

On Sun, Sep 8, 2019 at 2:52 AM Yves Hilpisch ***@***.***> wrote: I got Tensorflow 2. Anyone has ideas on how to fix to work with TF 2? Dont want to downgrade to TF 1.14. Thanks. What issues do you face? Tensorflow is used only indirectly via Keras. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <https://gist.github.com/08167918a51fc2adc8f20a92364886db?email_source=notifications&email_token=AEOC6YML6YQM4N5LWYBG5T3QISOKLA5CNFSM4IURYTT2YY3PNVWWK3TUL52HS4DFVNDWS43UINXW23LFNZ2KUY3PNVWWK3TUL5UWJTQAFYJ6K#gistcomment-3019749>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEOC6YPPYCNFCUFWNLWNTSDQISOKLANCNFSM4IURYTTQ> .

This has probably something to do with inconsistent versions overall.

Would try to upgrade all relevant packages (keras, tf, etc.)

That will be greatly appreciated. Also, regarding related courses, please send info. Thanks a lot.

…

On Sun, Sep 8, 2019 at 11:32 AM Yves Hilpisch ***@***.***> wrote: This has probably something to do with inconsistent versions overall. Would try to upgrade all relevant packages (keras, tf, etc.) — You are receiving this because you commented. Reply to this email directly, view it on GitHub <https://gist.github.com/08167918a51fc2adc8f20a92364886db?email_source=notifications&email_token=AEOC6YOQXUTHD6RDUOPDJ73QIULIZA5CNFSM4IURYTT2YY3PNVWWK3TUL52HS4DFVNDWS43UINXW23LFNZ2KUY3PNVWWK3TUL5UWJTQAFYKIS#gistcomment-3019913>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEOC6YPR3XLQLKD2CW4ZIU3QIULIZANCNFSM4IURYTTQ> .

I meant that you should try to upgrade all of your packages.

Have a look eg here: http://certificate.tpq.io

This is our most comprehensive offering in Python for Algorithmic Trading.

Thanks.

…

On Sun, Sep 8, 2019 at 11:39 AM Yves Hilpisch ***@***.***> wrote: I meant that you should try to upgrade all of your packages. Have a look eg here: http://certificate.tpq.io This is our most comprehensive offering in Python for Algorithmic Trading. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <https://gist.github.com/08167918a51fc2adc8f20a92364886db?email_source=notifications&email_token=AEOC6YJGXO76CLPZKDXBWQLQIUMC3A5CNFSM4IURYTT2YY3PNVWWK3TUL52HS4DFVNDWS43UINXW23LFNZ2KUY3PNVWWK3TUL5UWJTQAFYKI4#gistcomment-3019918>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEOC6YKIABOVMVXW3XYPS7LQIUMC3ANCNFSM4IURYTTQ> .