Comparing Results.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "90d79161",
   "metadata": {},
   "source": [
    "### Initialization process"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "c4f7e22e",
   "metadata": {},
   "outputs": [],
   "source": [
    "from sqlite3 import connect\n",
    "from pandas import DataFrame, Timedelta, concat, Series\n",
    "from ccxt import binance\n",
    "from datetime import datetime\n",
    "from collections import defaultdict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "6df33870",
   "metadata": {},
   "outputs": [],
   "source": [
    "conn = connect(\".\\cryptobot9.db\")\n",
    "cursor = conn.cursor()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fd91629f",
   "metadata": {},
   "source": [
    "### Data preparation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "02cc4a20",
   "metadata": {},
   "outputs": [],
   "source": [
    "ans = cursor.execute(\"SELECT * FROM indexes\")\n",
    "rows = ans.fetchall()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e1065fc8",
   "metadata": {},
   "source": [
    "tweet_id, user_id, sentments, sent_score, favorited, retweets, created_at, text, hashtags, mentions, urls, type, retweet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "cc705e1c",
   "metadata": {},
   "outputs": [],
   "source": [
    "df = DataFrame(rows, columns=[\"index_id\", \"forecast\", \"cryptocurrency\", \"timestamp\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "20089f71",
   "metadata": {},
   "outputs": [],
   "source": [
    "df.sort_values(by=\"timestamp\", inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "240ac573",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_btc = df.loc[df[\"cryptocurrency\"] == \"bitcoin\"]\n",
    "df_eth = df.loc[df[\"cryptocurrency\"] == \"ethereum\"]\n",
    "df_bnb = df.loc[df[\"cryptocurrency\"] == \"binance coin\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "c40f6a6d",
   "metadata": {},
   "outputs": [],
   "source": [
    "binance_btc = binance().fetchOHLCV(\"BTCUSDT\", timeframe = \"1h\")\n",
    "binance_eth = binance().fetchOHLCV(\"ETHUSDT\", timeframe = \"1h\")\n",
    "binance_bnb = binance().fetchOHLCV(\"BNBUSDT\", timeframe = \"1h\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "64e37c4e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_10202/287363170.py:1: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df_btc[\"datetime\"] = df_btc[\"timestamp\"].apply(lambda x: datetime.fromtimestamp(int(x)).replace(minute=0, second=0))\n",
      "/tmp/ipykernel_10202/287363170.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df_eth[\"datetime\"] = df_eth[\"timestamp\"].apply(lambda x: datetime.fromtimestamp(int(x)).replace(minute=0, second=0))\n",
      "/tmp/ipykernel_10202/287363170.py:3: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  df_bnb[\"datetime\"] = df_bnb[\"timestamp\"].apply(lambda x: datetime.fromtimestamp(int(x)).replace(minute=0, second=0))\n"
     ]
    }
   ],
   "source": [
    "df_btc[\"datetime\"] = df_btc[\"timestamp\"].apply(lambda x: datetime.fromtimestamp(int(x)).replace(minute=0, second=0))\n",
    "df_eth[\"datetime\"] = df_eth[\"timestamp\"].apply(lambda x: datetime.fromtimestamp(int(x)).replace(minute=0, second=0))\n",
    "df_bnb[\"datetime\"] = df_bnb[\"timestamp\"].apply(lambda x: datetime.fromtimestamp(int(x)).replace(minute=0, second=0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "63bae9cc",
   "metadata": {},
   "outputs": [],
   "source": [
    "bin_df_btc = DataFrame(binance_btc, columns=[\"timestamp\", \"open\", \"high\", \"low\", \"close\", \"volume\"])\n",
    "bin_df_btc[\"datetime\"] = bin_df_btc[\"timestamp\"].apply(lambda x: datetime.fromtimestamp(x/1000))\n",
    "bin_df_eth = DataFrame(binance_eth, columns=[\"timestamp\", \"open\", \"high\", \"low\", \"close\", \"volume\"])\n",
    "bin_df_eth[\"datetime\"] = bin_df_eth[\"timestamp\"].apply(lambda x: datetime.fromtimestamp(x/1000))\n",
    "bin_df_bnb = DataFrame(binance_bnb, columns=[\"timestamp\", \"open\", \"high\", \"low\", \"close\", \"volume\"])\n",
    "bin_df_bnb[\"datetime\"] = bin_df_bnb[\"timestamp\"].apply(lambda x: datetime.fromtimestamp(x/1000))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "76066462",
   "metadata": {},
   "source": [
    "### Dataframe Bitcoin per time distance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "0c638a2c",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_btc_now = df_btc.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_btc_1h = df_btc.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_btc_1h[\"datetime\"] += Timedelta(1, unit=\"H\")\n",
    "df_btc_2h = df_btc.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_btc_2h[\"datetime\"] += Timedelta(2, unit=\"H\")\n",
    "df_btc_3h = df_btc.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_btc_3h[\"datetime\"] += Timedelta(3, unit=\"H\")\n",
    "df_btc_6h = df_btc.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_btc_6h[\"datetime\"] += Timedelta(6, unit=\"H\")\n",
    "df_btc_12h = df_btc.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_btc_12h[\"datetime\"] += Timedelta(12, unit=\"H\")\n",
    "df_btc_24h = df_btc.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_btc_24h[\"datetime\"] += Timedelta(1, unit=\"D\")\n",
    "df_btc_48h = df_btc.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_btc_48h[\"datetime\"] += Timedelta(2, unit=\"D\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "c3dc9c1f",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_10202/4000265616.py:7: FutureWarning: The default dtype for empty Series will be 'object' instead of 'float64' in a future version. Specify a dtype explicitly to silence this warning.\n",
      "  df_btc_12h[\"open\"] = concat([Series([0]*(len(df_btc_12h)-len(temp_df))), temp_df]).values\n"
     ]
    }
   ],
   "source": [
    "df_btc_now[\"open\"] = bin_df_btc.loc[bin_df_btc[\"datetime\"].isin(df_btc_now[\"datetime\"])][\"open\"].values\n",
    "df_btc_1h[\"open\"] = bin_df_btc.loc[bin_df_btc[\"datetime\"].isin(df_btc_1h[\"datetime\"])][\"open\"].values\n",
    "df_btc_2h[\"open\"] = bin_df_btc.loc[bin_df_btc[\"datetime\"].isin(df_btc_2h[\"datetime\"])][\"open\"].values\n",
    "df_btc_3h[\"open\"] = bin_df_btc.loc[bin_df_btc[\"datetime\"].isin(df_btc_3h[\"datetime\"])][\"open\"].values\n",
    "df_btc_6h[\"open\"] = bin_df_btc.loc[bin_df_btc[\"datetime\"].isin(df_btc_6h[\"datetime\"])][\"open\"].values\n",
    "temp_df = bin_df_btc.loc[bin_df_btc[\"datetime\"].isin(df_btc_12h[\"datetime\"])][\"open\"]\n",
    "df_btc_12h[\"open\"] = concat([Series([0]*(len(df_btc_12h)-len(temp_df))), temp_df]).values\n",
    "temp_df = bin_df_btc.loc[bin_df_btc[\"datetime\"].isin(df_btc_24h[\"datetime\"])][\"open\"]\n",
    "df_btc_24h[\"open\"] = concat([Series([0]*(len(df_btc_24h)-len(temp_df))), temp_df]).values\n",
    "temp_df = bin_df_btc.loc[bin_df_btc[\"datetime\"].isin(df_btc_48h[\"datetime\"])][\"open\"]\n",
    "df_btc_48h[\"open\"] = concat([Series([0]*(len(df_btc_48h)-len(temp_df))), temp_df]).values"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a6b30af2",
   "metadata": {},
   "source": [
    "### Dataframe Ethereum per time distance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "bf74e1a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_eth_now = df_eth.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_eth_1h = df_eth.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_eth_1h[\"datetime\"] += Timedelta(1, unit=\"H\")\n",
    "df_eth_2h = df_eth.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_eth_2h[\"datetime\"] += Timedelta(2, unit=\"H\")\n",
    "df_eth_3h = df_eth.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_eth_3h[\"datetime\"] += Timedelta(3, unit=\"H\")\n",
    "df_eth_6h = df_eth.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_eth_6h[\"datetime\"] += Timedelta(6, unit=\"H\")\n",
    "df_eth_12h = df_eth.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_eth_12h[\"datetime\"] += Timedelta(12, unit=\"H\")\n",
    "df_eth_24h = df_eth.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_eth_24h[\"datetime\"] += Timedelta(1, unit=\"D\")\n",
    "df_eth_48h = df_eth.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_eth_48h[\"datetime\"] += Timedelta(2, unit=\"D\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "403d6a1a",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_10202/2258726641.py:7: FutureWarning: The default dtype for empty Series will be 'object' instead of 'float64' in a future version. Specify a dtype explicitly to silence this warning.\n",
      "  df_eth_12h[\"open\"] = concat([Series([0]*(len(df_eth_12h)-len(temp_df))), temp_df]).values\n"
     ]
    }
   ],
   "source": [
    "df_eth_now[\"open\"] = bin_df_eth.loc[bin_df_eth[\"datetime\"].isin(df_eth_now[\"datetime\"])][\"open\"].values\n",
    "df_eth_1h[\"open\"] = bin_df_eth.loc[bin_df_eth[\"datetime\"].isin(df_eth_1h[\"datetime\"])][\"open\"].values\n",
    "df_eth_2h[\"open\"] = bin_df_eth.loc[bin_df_eth[\"datetime\"].isin(df_eth_2h[\"datetime\"])][\"open\"].values\n",
    "df_eth_3h[\"open\"] = bin_df_eth.loc[bin_df_eth[\"datetime\"].isin(df_eth_3h[\"datetime\"])][\"open\"].values\n",
    "df_eth_6h[\"open\"] = bin_df_eth.loc[bin_df_eth[\"datetime\"].isin(df_eth_6h[\"datetime\"])][\"open\"].values\n",
    "temp_df = bin_df_eth.loc[bin_df_eth[\"datetime\"].isin(df_eth_12h[\"datetime\"])][\"open\"]\n",
    "df_eth_12h[\"open\"] = concat([Series([0]*(len(df_eth_12h)-len(temp_df))), temp_df]).values\n",
    "temp_df = bin_df_eth.loc[bin_df_eth[\"datetime\"].isin(df_eth_24h[\"datetime\"])][\"open\"]\n",
    "df_eth_24h[\"open\"] = concat([Series([0]*(len(df_eth_24h)-len(temp_df))), temp_df]).values\n",
    "temp_df = bin_df_eth.loc[bin_df_eth[\"datetime\"].isin(df_eth_48h[\"datetime\"])][\"open\"]\n",
    "df_eth_48h[\"open\"] = concat([Series([0]*(len(df_eth_48h)-len(temp_df))), temp_df]).values"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cdcb4d20",
   "metadata": {},
   "source": [
    "### Dataframe Ethereum per time distance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "b276df2b",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_bnb_now = df_bnb.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_bnb_1h = df_bnb.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_bnb_1h[\"datetime\"] += Timedelta(1, unit=\"H\")\n",
    "df_bnb_2h = df_bnb.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_bnb_2h[\"datetime\"] += Timedelta(2, unit=\"H\")\n",
    "df_bnb_3h = df_bnb.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_bnb_3h[\"datetime\"] += Timedelta(3, unit=\"H\")\n",
    "df_bnb_6h = df_bnb.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_bnb_6h[\"datetime\"] += Timedelta(6, unit=\"H\")\n",
    "df_bnb_12h = df_bnb.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_bnb_12h[\"datetime\"] += Timedelta(12, unit=\"H\")\n",
    "df_bnb_24h = df_bnb.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_bnb_24h[\"datetime\"] += Timedelta(1, unit=\"D\")\n",
    "df_bnb_48h = df_bnb.groupby(by=\"datetime\")[\"forecast\"].mean().reset_index()\n",
    "df_bnb_48h[\"datetime\"] += Timedelta(2, unit=\"D\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "ceb660b6",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/tmp/ipykernel_10202/484217101.py:7: FutureWarning: The default dtype for empty Series will be 'object' instead of 'float64' in a future version. Specify a dtype explicitly to silence this warning.\n",
      "  df_bnb_12h[\"open\"] = concat([Series([0]*(len(df_bnb_12h)-len(temp_df))), temp_df]).values\n"
     ]
    }
   ],
   "source": [
    "df_bnb_now[\"open\"] = bin_df_bnb.loc[bin_df_bnb[\"datetime\"].isin(df_bnb_now[\"datetime\"])][\"open\"].values\n",
    "df_bnb_1h[\"open\"] = bin_df_bnb.loc[bin_df_bnb[\"datetime\"].isin(df_bnb_1h[\"datetime\"])][\"open\"].values\n",
    "df_bnb_2h[\"open\"] = bin_df_bnb.loc[bin_df_bnb[\"datetime\"].isin(df_bnb_2h[\"datetime\"])][\"open\"].values\n",
    "df_bnb_3h[\"open\"] = bin_df_bnb.loc[bin_df_bnb[\"datetime\"].isin(df_bnb_3h[\"datetime\"])][\"open\"].values\n",
    "df_bnb_6h[\"open\"] = bin_df_bnb.loc[bin_df_bnb[\"datetime\"].isin(df_bnb_6h[\"datetime\"])][\"open\"].values\n",
    "temp_df = bin_df_bnb.loc[bin_df_bnb[\"datetime\"].isin(df_bnb_12h[\"datetime\"])][\"open\"]\n",
    "df_bnb_12h[\"open\"] = concat([Series([0]*(len(df_bnb_12h)-len(temp_df))), temp_df]).values\n",
    "temp_df = bin_df_bnb.loc[bin_df_bnb[\"datetime\"].isin(df_bnb_24h[\"datetime\"])][\"open\"]\n",
    "df_bnb_24h[\"open\"] = concat([Series([0]*(len(df_bnb_24h)-len(temp_df))), temp_df]).values\n",
    "temp_df = bin_df_bnb.loc[bin_df_bnb[\"datetime\"].isin(df_bnb_48h[\"datetime\"])][\"open\"]\n",
    "df_bnb_48h[\"open\"] = concat([Series([0]*(len(df_bnb_48h)-len(temp_df))), temp_df]).values"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "45e4f97b",
   "metadata": {},
   "source": [
    "### Correlation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "2e041e93",
   "metadata": {},
   "outputs": [],
   "source": [
    "corr = {\n",
    "    \"btc_0h\":df_btc_now[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"btc_1h\":df_btc_1h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"btc_2h\":df_btc_2h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"btc_3h\":df_btc_3h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"btc_6h\":df_btc_6h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"btc_12h\":df_btc_12h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"btc_1d\":df_btc_24h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"btc_2d\":df_btc_48h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"eth_0h\":df_eth_now[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"eth_1h\":df_eth_1h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"eth_2h\":df_eth_2h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"eth_3h\":df_eth_3h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"eth_6h\":df_eth_6h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"eth_12h\":df_eth_12h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"eth_1d\":df_eth_24h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"eth_2d\":df_eth_48h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"bnb_0h\":df_bnb_now[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"bnb_1h\":df_bnb_1h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"bnb_2h\":df_bnb_2h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"bnb_3h\":df_bnb_3h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"bnb_6h\":df_bnb_6h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"bnb_12h\":df_bnb_12h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"bnb_1d\":df_bnb_24h[[\"forecast\", \"open\"]].corr()[\"open\"][0],\n",
    "    \"bnb_2d\":df_bnb_48h[[\"forecast\", \"open\"]].corr()[\"open\"][0]\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "5c7afbc2",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'btc_0h': 0.15172086456558656,\n",
       " 'btc_1h': -0.24873156700973148,\n",
       " 'btc_2h': -0.07687721401842587,\n",
       " 'btc_3h': -0.012092985357873634,\n",
       " 'btc_6h': -0.020408788315175244,\n",
       " 'btc_12h': -0.010054329553099364,\n",
       " 'btc_1d': 0.19561644879226803,\n",
       " 'btc_2d': nan,\n",
       " 'eth_0h': 0.10579726481084044,\n",
       " 'eth_1h': 0.028868850189259266,\n",
       " 'eth_2h': 0.03353439872681746,\n",
       " 'eth_3h': 0.014580478724054638,\n",
       " 'eth_6h': -0.7206644523878813,\n",
       " 'eth_12h': 0.5405765545427734,\n",
       " 'eth_1d': -0.4586160756517969,\n",
       " 'eth_2d': nan,\n",
       " 'bnb_0h': -0.3386812109396907,\n",
       " 'bnb_1h': 0.030194576528928225,\n",
       " 'bnb_2h': -0.18700647257649008,\n",
       " 'bnb_3h': -0.4633012073704392,\n",
       " 'bnb_6h': 0.01799341969234186,\n",
       " 'bnb_12h': 0.08783150528073844,\n",
       " 'bnb_1d': -0.6498469642033794,\n",
       " 'bnb_2d': nan}"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e488691c",
   "metadata": {},
   "source": [
    "### Calculating the percentual forecast"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "f688f4c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "forecast_price = lambda p, tv, t: p + tv if t else tv - p"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "25c9b6e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "df_btc_now[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_btc_now[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_btc_1h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_btc_1h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_btc_2h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_btc_2h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_btc_3h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_btc_3h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_btc_6h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_btc_6h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_btc_12h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_btc_12h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_btc_24h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_btc_24h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_btc_48h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_btc_48h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_eth_now[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_eth_now[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_eth_1h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_eth_1h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_eth_2h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_eth_2h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_eth_3h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_eth_3h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_eth_6h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_eth_6h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_eth_12h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_eth_12h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_eth_24h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_eth_24h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_eth_48h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_eth_48h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_bnb_now[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_bnb_now[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_bnb_1h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_bnb_1h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_bnb_2h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_bnb_2h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_bnb_3h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_bnb_3h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_bnb_6h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_bnb_6h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_bnb_12h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_bnb_12h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_bnb_24h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_bnb_24h[[\"forecast\", \"open\"]].values][:-1]\n",
    "df_bnb_48h[\"price_pred\"] = [0] + [forecast_price(o*(abs(f)/1000), o, f>=0) for f, o in df_bnb_48h[[\"forecast\", \"open\"]].values][:-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d032fe8e",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "01072205",
   "metadata": {},
   "source": [
    "### Calculating precision, accuracy, recall and, f-1 score"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "e0d3d315",
   "metadata": {},
   "outputs": [],
   "source": [
    "def performance_check(forecast, price):\n",
    "    if price == 0 or forecast == 0:\n",
    "        return \"ND\"\n",
    "    elif forecast > 0 and price > 0:\n",
    "        return \"TP\"\n",
    "    elif forecast < 0 and price < 0:\n",
    "        return \"TN\"\n",
    "    elif forecast > 0 and price < 0:\n",
    "        return \"FP\"\n",
    "    elif forecast < 0 and price > 0:\n",
    "        return \"FN\"\n",
    "    else:\n",
    "        return \"ERR\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "id": "e61e1b53",
   "metadata": {},
   "outputs": [],
   "source": [
    "dfs_btc = {\n",
    "    \"df_btc_now\": df_btc_now, \n",
    "    \"df_btc_1h\": df_btc_1h, \n",
    "    \"df_btc_2h\": df_btc_2h, \n",
    "    \"df_btc_3h\": df_btc_3h, \n",
    "    \"df_btc_6h\": df_btc_6h, \n",
    "    \"df_btc_12h\": df_btc_12h, \n",
    "    \"df_btc_24h\": df_btc_24h \n",
    "#     \"df_btc_48h\": df_btc_48h\n",
    "}\n",
    "\n",
    "dfs_eth = {\n",
    "    \"df_eth_now\": df_eth_now, \n",
    "    \"df_eth_1h\": df_eth_1h, \n",
    "    \"df_eth_2h\": df_eth_2h, \n",
    "    \"df_eth_3h\": df_eth_3h, \n",
    "    \"df_eth_6h\": df_eth_6h, \n",
    "    \"df_eth_12h\": df_eth_12h, \n",
    "    \"df_eth_24h\": df_eth_24h\n",
    "#     \"df_eth_48h\": df_eth_48h\n",
    "}\n",
    "\n",
    "dfs_bnb = {\n",
    "    \"df_bnb_now\": df_bnb_now, \n",
    "    \"df_bnb_1h\": df_bnb_1h, \n",
    "    \"df_bnb_2h\": df_bnb_2h, \n",
    "    \"df_bnb_3h\": df_bnb_3h, \n",
    "    \"df_bnb_6h\": df_bnb_6h, \n",
    "    \"df_bnb_12h\": df_bnb_12h, \n",
    "    \"df_bnb_24h\": df_bnb_24h\n",
    "#     \"df_bnb_48h\": df_bnb_48h\n",
    "}\n",
    "dfs = [dfs_btc, dfs_eth, dfs_bnb]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "id": "384094fa",
   "metadata": {},
   "outputs": [],
   "source": [
    "performance = {}\n",
    "for each_df in dfs:\n",
    "    for df_name, df in each_df.items():\n",
    "        results = [\"ND\"]\n",
    "        performance[df_name] = {}\n",
    "        for step in range(1, len(df)):\n",
    "            results.append(performance_check(df[\"forecast\"][step-1], df[\"open\"][step] - df[\"open\"][step-1]))\n",
    "        df[\"performance\"] = results\n",
    "        temp = defaultdict(lambda: 0, df[\"performance\"].value_counts().to_dict())\n",
    "        performance[df_name] = {\n",
    "            \"accuracy\" : temp[\"TP\"] + temp[\"TN\"] / sum(temp.values()),\n",
    "            \"preciosion\": temp[\"TP\"] / (temp[\"TP\"] + temp[\"FP\"]),\n",
    "            \"recall\": temp[\"TP\"] / (temp[\"TP\"] + temp[\"FN\"]),\n",
    "            \"f1-score\": temp[\"TP\"] / (temp[\"TP\"] + 1/2 * (temp[\"FP\"] + temp[\"FN\"]))\n",
    "        }"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "id": "888fb2d9",
   "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>datetime</th>\n",
       "      <th>forecast</th>\n",
       "      <th>open</th>\n",
       "      <th>price_pred</th>\n",
       "      <th>performance</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2022-08-23 09:00:00</td>\n",
       "      <td>7.212758</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2022-08-23 10:00:00</td>\n",
       "      <td>6.589048</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2022-08-23 11:00:00</td>\n",
       "      <td>15.819305</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2022-08-23 12:00:00</td>\n",
       "      <td>40.261843</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2022-08-23 13:00:00</td>\n",
       "      <td>9.035561</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2022-08-23 14:00:00</td>\n",
       "      <td>8.009561</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2022-08-23 15:00:00</td>\n",
       "      <td>1.949842</td>\n",
       "      <td>298.7</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>TP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>2022-08-23 16:00:00</td>\n",
       "      <td>0.810513</td>\n",
       "      <td>296.5</td>\n",
       "      <td>299.282418</td>\n",
       "      <td>FP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>2022-08-23 17:00:00</td>\n",
       "      <td>0.279415</td>\n",
       "      <td>298.3</td>\n",
       "      <td>296.740317</td>\n",
       "      <td>TP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>2022-08-23 18:00:00</td>\n",
       "      <td>0.672027</td>\n",
       "      <td>298.1</td>\n",
       "      <td>298.383350</td>\n",
       "      <td>FP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>2022-08-23 19:00:00</td>\n",
       "      <td>2.947457</td>\n",
       "      <td>298.2</td>\n",
       "      <td>298.300331</td>\n",
       "      <td>TP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2022-08-23 20:00:00</td>\n",
       "      <td>0.230815</td>\n",
       "      <td>297.2</td>\n",
       "      <td>299.078932</td>\n",
       "      <td>FP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>2022-08-23 21:00:00</td>\n",
       "      <td>0.151538</td>\n",
       "      <td>296.9</td>\n",
       "      <td>297.268598</td>\n",
       "      <td>FP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>2022-08-23 22:00:00</td>\n",
       "      <td>0.316938</td>\n",
       "      <td>298.4</td>\n",
       "      <td>296.944992</td>\n",
       "      <td>TP</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>2022-08-23 23:00:00</td>\n",
       "      <td>1.674044</td>\n",
       "      <td>298.4</td>\n",
       "      <td>298.494574</td>\n",
       "      <td>ND</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>2022-08-24 00:00:00</td>\n",
       "      <td>5.123852</td>\n",
       "      <td>298.0</td>\n",
       "      <td>298.899535</td>\n",
       "      <td>FP</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              datetime   forecast   open  price_pred performance\n",
       "0  2022-08-23 09:00:00   7.212758    0.0    0.000000          ND\n",
       "1  2022-08-23 10:00:00   6.589048    0.0    0.000000          ND\n",
       "2  2022-08-23 11:00:00  15.819305    0.0    0.000000          ND\n",
       "3  2022-08-23 12:00:00  40.261843    0.0    0.000000          ND\n",
       "4  2022-08-23 13:00:00   9.035561    0.0    0.000000          ND\n",
       "5  2022-08-23 14:00:00   8.009561    0.0    0.000000          ND\n",
       "6  2022-08-23 15:00:00   1.949842  298.7    0.000000          TP\n",
       "7  2022-08-23 16:00:00   0.810513  296.5  299.282418          FP\n",
       "8  2022-08-23 17:00:00   0.279415  298.3  296.740317          TP\n",
       "9  2022-08-23 18:00:00   0.672027  298.1  298.383350          FP\n",
       "10 2022-08-23 19:00:00   2.947457  298.2  298.300331          TP\n",
       "11 2022-08-23 20:00:00   0.230815  297.2  299.078932          FP\n",
       "12 2022-08-23 21:00:00   0.151538  296.9  297.268598          FP\n",
       "13 2022-08-23 22:00:00   0.316938  298.4  296.944992          TP\n",
       "14 2022-08-23 23:00:00   1.674044  298.4  298.494574          ND\n",
       "15 2022-08-24 00:00:00   5.123852  298.0  298.899535          FP"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_bnb_24h"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "id": "10a71ea0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'precision')"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot\n",
    "\n",
    "# btc\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"accuracy\"] for key, value in performance.items() if \"btc\" in key],\n",
    "            label=\"btc\"\n",
    "           )\n",
    "# eth\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"accuracy\"] for key, value in performance.items() if \"eth\" in key],\n",
    "            label=\"eth\"\n",
    "           )\n",
    "# bnb\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"accuracy\"] for key, value in performance.items() if \"bnb\" in key],\n",
    "            label=\"bnb\"\n",
    "           )\n",
    "\n",
    "pyplot.xlabel(\"Time Distance\")\n",
    "pyplot.ylabel(\"accuracy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "id": "8bf55438",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'precision')"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot\n",
    "\n",
    "# btc\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"preciosion\"] for key, value in performance.items() if \"btc\" in key],\n",
    "            label=\"btc\"\n",
    "           )\n",
    "# eth\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"preciosion\"] for key, value in performance.items() if \"eth\" in key],\n",
    "            label=\"eth\"\n",
    "           )\n",
    "# bnb\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"preciosion\"] for key, value in performance.items() if \"bnb\" in key],\n",
    "            label=\"bnb\"\n",
    "           )\n",
    "\n",
    "pyplot.xlabel(\"Time Distance\")\n",
    "pyplot.ylabel(\"precision\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 95,
   "id": "9de58365",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'precision')"
      ]
     },
     "execution_count": 95,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEGCAYAAAB/+QKOAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAAg/UlEQVR4nO3deZRcZ33m8e+jbu375k2t3TZG2MaLLCQTYgcmRJCMPZgE2xCCIRMnEyCQg09iT3JsxoTDQDyZDAHCcYgxDgzEIyDHBMXG8QLMdAnUXiRZMjKqtmW1ZFvd1Vqstbff/FG37XK71F1q9e1bVf18zumjuu+9VfW7kqqfuve9930VEZiZmQ00LusCzMysOjkgzMysLAeEmZmV5YAwM7OyHBBmZlZWY9YFjJR58+bFkiVLsi7DzKymPPbYYx0RMb/curoJiCVLltDS0pJ1GWZmNUXSzhOt8ykmMzMrywFhZmZlOSDMzKwsB4SZmZXlgDAzs7JSCwhJd0naK+mpE6yXpC9K2iFps6RLStZ9SNIvk58PpVWjmZmdWJpHEHcDawdZ/y7gnOTnRuDvASTNAW4D3gKsAm6TNDvFOs3MrIzU7oOIiJ9IWjLIJlcD90RxvPENkmZJOhO4EngwIjoBJD1IMWi+nVatn7/3P/KLIy+k9fKjJgi6evqyLsPMRtnCxtP5/A33j/jrZnmj3AJgV8lyW9J2ovbXkXQjxaMPFi1alE6VNaS7JzjugDAbc/Z3daXyujV9J3VE3AncCbBy5cphz3z05+/7wYjVlKWP3L2RFw4c5uFPXZl1KWZWB7K8imk3sLBkuSlpO1G7DaK7t4+ftRZYs2xu1qWYWZ3IMiDuA34vuZppNXAgIl4AHgDeKWl20jn9zqTNBrFl9wEOd/Vy+fJ5WZdiZnUitVNMkr5NscN5nqQ2ilcmjQeIiK8C64F3AzuAI8CHk3Wdkj4DbExe6vb+Dms7sVy+AMDqZXMyrsTM6kWaVzFdP8T6AD56gnV3AXelUVe9yuULnHfGdOZOm5h1KWZWJ3wndR043tPLxuc6WbPc/Q9mNnIcEHXgief3c7ynz/0PZjaiHBB1IJcvME6waqn7H8xs5Dgg6kAuX+D8BTOZOXl81qWYWR1xQNS4o129PLFrn/sfzGzEOSBqXMvOTrp7w/0PZjbiHBA1rjlfoHGcWLnYA96a2chyQNS45nyBixbOYurEmh5Wy8yqkAOihh081s2Wtv1c7v4HM0uBA6KGbXy2k76A1Q4IM0uBA6KGNecLTGgcxyWL3P9gZiPPAVHDcvkCKxfPZtL4hqxLMbM65ICoUfsOd7HthYPufzCz1DggatSG1uLw3r5BzszS4oCoUc35AlMmNHBh06ysSzGzOuWAqFG51gKrls5hfIP/Cc0sHf7tUoP2HjzGjr2HPP+0maXKAVGDckn/g8dfMrM0OSBqUC5fYMakRlacNSPrUsysjjkgalBzvsBbls2lYZyyLsXM6pgDosa07TvC851HfP+DmaXOAVFjcnn3P5jZ6HBA1JhcvsDcqRM49/RpWZdiZnUu1YCQtFbSdkk7JN1cZv1iSQ9J2izpUUlNJeu+IGmrpKclfVHSmD/hHhE05wusXj4X/3WYWdpSCwhJDcCXgXcBK4DrJa0YsNkdwD0RcSFwO/C55LmXA28FLgTOBy4Drkir1lrxbMdhXjx4zP0PZjYq0jyCWAXsiIjWiOgCvgNcPWCbFcDDyeNHStYHMAmYAEwExgMvpVhrTfD9D2Y2mtIMiAXArpLltqSt1CbgmuTxe4DpkuZGRI5iYLyQ/DwQEU8PfANJN0pqkdTS3t4+4jtQbZrzBc6YMYklc6dkXYqZjQFZd1LfBFwh6QmKp5B2A72SzgbeCDRRDJW3S3rbwCdHxJ0RsTIiVs6fP3806x51EcGGfIHL3f9gZqMkzZnudwMLS5abkrZXRMQekiMISdOA90bEfkl/AGyIiEPJun8D1gA/TbHeqvbMS4coHO7y8N5mNmrSPILYCJwjaamkCcB1wH2lG0iaJ6m/hluAu5LHz1M8smiUNJ7i0cXrTjGNJc35DsDzP5jZ6EktICKiB/gY8ADFX+73RsRWSbdLuirZ7Epgu6RngNOBzybt64A8sIViP8WmiPhBWrXWguZ8gUVzptA02/0PZjY60jzFRESsB9YPaLu15PE6imEw8Hm9wB+mWVst6e0LftZa4N0XnJl1KWY2hmTdSW0V2LbnIAeP9fj0kpmNKgdEDXil/8ETBJnZKHJA1IBca4GzT5vGaTMmZV2KmY0hDogq193bx8+f7fTwGmY26hwQVW5z236OdPX69JKZjToHRJVr3lEcf2m1A8LMRpkDosrlWgusOHMGs6dOyLoUMxtjHBBV7Fh3Ly079/nyVjPLhAOiij3+/D66evrcQW1mmXBAVLEN+QIN48SqpXOyLsXMxiAHRBVrzhc4f8FMpk8an3UpZjYGOSCq1OHjPTy5a79PL5lZZhwQVapl5z56+sIBYWaZcUBUqeZ8B+MbxMrF7n8ws2w4IKpULl/g4oWzmTyhIetSzGyMckBUoQNHu3lq9wHf/2BmmXJAVKGfP9tJX+D+BzPLlAOiCjXnO5jYOI6LFs3KuhQzG8McEFUoly9w2ZI5TGx0/4OZZccBUWUKh47zixdfdv+DmWXOAVFlNrR2AjggzCxzDogq05zvYNrERi5cMDPrUsxsjEs1ICStlbRd0g5JN5dZv1jSQ5I2S3pUUlPJukWSfiTpaUnbJC1Js9ZqkWstsGrpHBobnN1mlq3UfgtJagC+DLwLWAFcL2nFgM3uAO6JiAuB24HPlay7B/jriHgjsArYm1at1eLFA8dobT/s6UXNrCqk+TV1FbAjIlojogv4DnD1gG1WAA8njx/pX58ESWNEPAgQEYci4kiKtVaFXGsH4P4HM6sOaQbEAmBXyXJb0lZqE3BN8vg9wHRJc4Fzgf2SvifpCUl/nRyRvIakGyW1SGppb29PYRdGV/OOAjMnj2fFmTOyLsXMLPNO6puAKyQ9AVwB7AZ6gUbgbcn6y4BlwA0DnxwRd0bEyohYOX/+/FErOi251gJrls1l3DhlXYqZWaoBsRtYWLLclLS9IiL2RMQ1EXEx8BdJ236KRxtPJqeneoB/AS5JsdbM7eo8Qtu+oz69ZGZVI82A2AicI2mppAnAdcB9pRtImiepv4ZbgLtKnjtLUv9hwduBbSnWmrnmfLH/weMvmVm1SC0gkm/+HwMeAJ4G7o2IrZJul3RVstmVwHZJzwCnA59NnttL8fTSQ5K2AAL+Ia1aq0EuX2DetImcfdq0rEsxMwOK5/pTExHrgfUD2m4tebwOWHeC5z4IXJhmfdUiImjOF1izfC6S+x/MrDpk3UltQL79MHtfPu7TS2ZWVRwQVSDXWgDc/2Bm1cUBUQVy+Q7OmjmJRXOmZF2KmdkrHBAZ6+sLcvkCa5bPc/+DmVUVB0TGtr/0MvuOdPv0kplVHQdExprzxf4H3yBnZtXGAZGxXL6DJXOncNasyVmXYmb2Gg6IDPX09vGz1k7WLJ+XdSlmZq/jgMjQ1j0Hefl4j/sfzKwqOSAy1N//sNoTBJlZFap4qA1JC4DFpc+JiJ+kUdRY0Zzv4NzTpzF/+sSsSzEze52KAkLS54FrKY6o2ps0B+CAGKaunj5antvHtZctHHpjM7MMVHoE8Z+AN0TE8RRrGVM2te3naHevTy+ZWdWqtA+iFRifZiFjTfOOAhKsXjYn61LMzMqq9AjiCPCkpIeAV44iIuJPUqlqDMi1dvCms2Ywa8qErEsxMyur0oC4jwGzwdnwHevu5fGd+/nQ5YuzLsXM7IQqCoiI+EYybei5SdP2iOhOr6z69tjOfXT19nG5b5AzsypW6VVMVwLfAJ6jOP3nQkkf8mWuw9Oc76BhnLhsqfsfzKx6VXqK6X8A74yI7QCSzgW+DVyaVmH1LJcv8OammUybmOqMr2Zmp6TSq5jG94cDQEQ8g69qGpZDx3vY1HbAo7eaWdWr9Ctsi6SvAd9Mlj8AtKRTUn3b+GwnvX3h/gczq3qVBsR/AT4K9F/W+lPgK6lUVOdyrQUmNIzj0sWzsy7FzGxQlV7FdBz4m+THTkFzvoOLF81i0viGrEsxMxvUoH0Qku5N/twiafPAn6FeXNJaSdsl7ZB0c5n1iyU9lLzeo5KaBqyfIalN0pdOdseq0f4jXWzdc9Cnl8ysJgx1BPGJ5M/fOtkXltQAfBn4daAN2CjpvojYVrLZHcA9yX0Wbwc+B3ywZP1nqKMBAX/2bCcRcPnZ7qA2s+o36BFERLyQPOwAdkXETmAi8GZgzxCvvQrYERGtEdEFfAe4esA2K4CHk8ePlK6XdClwOvCjCvajJuTyBSaPb+DNTbOyLsXMbEiVXub6E2BSMifEjyh+y797iOcsAHaVLLclbaU2Adckj98DTJc0V9I4ivde3DTYG0i6UVKLpJb29vaKdiRLzfkOVi6ZzYRGz9NkZtWv0t9UiogjFH+ZfyUifgd40wi8/03AFZKeAK4AdlOcb+KPgfUR0TbYkyPizohYGREr58+fPwLlpKf95eM889Ih9z+YWc2o9DJXSVpD8f6H30/ahroMZzdQOhtOU9L2iojYQ3IEIWka8N6I2J+819sk/TEwDZgg6VBEvK6ju1ZsaC1OL+r5p82sVlQaEJ8EbgG+HxFbJS2j2GcwmI3AOZKWUgyG64D3l24gaR7QGRF9yevfBRARHyjZ5gZgZS2HAxTnn54+sZE3nTUj61LMzCpS6X0QPwZ+XLLcyqs3zZ3oOT2SPgY8QPFo464kXG4HWiLiPuBK4HOS+qcv/eiw9qIG5PIdvGXZHBob3P9gZrVh0ICQ9LcR8UlJP6A4B/VrRMRVgz0/ItYD6we03VryeB2wbojXuJuhO8Sr2p79R3mucIQPrlmSdSlmZhUb6gjin5I/70i7kHqWyxf7H9Z4/mkzqyGDBkREPJY8bAGOJn0F/TfBTUy5trrRnC8we8p4zjtjetalmJlVrNIT4g8BU0qWJwP/PvLl1J+IYENrgTXL5zJunLIux8ysYpUGxKSIONS/kDyeMsj2lni+8wi79x/16SUzqzmVBsRhSZf0LyTDYBxNp6T60tzf/+Ab5MysxpzMfRD/R9IeinNSnwFcm1ZR9SSXL3Da9Iksnz8161LMzE5KpfdBbJR0HvCGpGl7RHSnV1Z9iAia8wXeevZcJPc/mFltqegUk6QpwJ8Dn4iIp4Alkk56CPCxZsfeQ3QcOu7hNcysJlXaB/F1oAtYkyzvBv4qlYrqSH//gwfoM7NaVGlALI+ILwDdAMnIrj5nMoRcvkDT7MksnOMLvsys9lQaEF2SJpMMtyFpOXA8tarqQF9fkGst+PJWM6tZlV7FdBtwP7BQ0reAtwI3pFVUPdj2wkEOHO329KJmVrOGDIhkdrfZFOdtWE3x1NInIqIj5dpqWv/8D2uWuf/BzGrTkAEREX2S/iwi7gV+OAo11YXmfIFl86ZyxsxJWZdiZjYslfZB/LukmyQtlDSn/yfVympYT28fP3+2kzW+vNXMalilfRDXUuyg/uMB7ctGtpz6sGX3AQ4d7/HlrWZW0yoNiBUUw+FXKAbFT4GvplVUreu//2H1Mh9kmVntqjQgvgEcBL6YLL8/aXtfGkXVuly+wHlnTGfuNE+ZYWa1q9KAOD8iVpQsPyJpWxoF1brjPb207Ozk+lWLsi7FzOyUVNpJ/bik1f0Lkt5CcZY5G+DJ5/dzrLvP/Q9mVvMqPYK4FGiW9HyyvAjYLmkLEBFxYSrV1aDmfIFxglVL3f9gZrWt0oBYm2oVdSSXL3D+gpnMnDw+61LMzE5JpfNB7Ey7kHpwtKuXJ3bt4yO/sjTrUszMTlmlfRDDImmtpO2Sdki6ucz6xZIekrRZ0qOSmpL2iyTlJG1N1tXE7HUtOzvp7g0P0GdmdSG1gJDUAHwZeBfF+yiul7RiwGZ3APckfRi3A59L2o8AvxcRb6J4eutvJc1Kq9aR0pwv0DhOXLbE/Q9mVvvSPIJYBeyIiNaI6AK+A1w9YJsVwMPJ40f610fEMxHxy+TxHmAvMD/FWkdELl/gooWzmDqx0q4dM7PqlWZALAB2lSy3JW2lNlEcJRbgPcB0Sa85PyNpFTAByA98A0k3SmqR1NLe3j5ihQ/HwWPdbG7b7/GXzKxupNoHUYGbgCskPQFcQXEq097+lZLOBP4J+HBE9A18ckTcGRErI2Ll/PnZHmBsfLaTvsABYWZ1I81zIbuBhSXLTUnbK5LTR9cASJoGvDci9ifLMygOL/4XEbEhxTpHRC5fYELjOC5ZNDvrUszMRkSaRxAbgXMkLZU0AbgOuK90A0nzkgmJAG4B7kraJwDfp9iBvS7FGkdMc77ApYtmM2l8Q9almJmNiNQCIiJ6gI8BDwBPA/dGxFZJt0u6KtnsSop3ZD8DnA58Nml/H/CrwA2Snkx+Lkqr1lO173AX2144yOU+vWRmdSTVy20iYj2wfkDbrSWP1wGvO0KIiG8C30yztpHUP72o5582s3qSdSd1Xci1FpgyoYELm2ZlXYqZ2YhxQIyA5nyBy5bMYXyD/zrNrH74N9op2nvwGDv2HnL/g5nVHQfEKcr19z94/gczqzMOiFOUyxeYMamRFWfNyLoUM7MR5YA4Rc35Am9ZNpeGccq6FDOzEeWAOAVt+47wfOcR9z+YWV1yQJyCXL7Y/+Dxl8ysHjkgTkEuX2Du1Amce9r0rEsxMxtxDohhighyrQVWL5/LOPc/mFkdckAM03OFI7xw4Jj7H8ysbjkghqk53wHg+afNrG45IIapOV/gjBmTWDpvatalmJmlwgExDBHBhnyBy5fPRXL/g5nVJwfEMDzz0iEKh7tY7f4HM6tjDohh6O9/cAe1mdUzB8Qw5PIFFs2ZQtPsKVmXYmaWGgfESertCza0Fnz1kpnVPQfESdq25yAHj/V4elEzq3sOiJOUa/X9D2Y2NjggTlJzvsDy+VM5bcakrEsxM0uVA+IkdPf28fNnOz17nJmNCakGhKS1krZL2iHp5jLrF0t6SNJmSY9KaipZ9yFJv0x+PpRmnZXa3LafI129vrzVzMaE1AJCUgPwZeBdwArgekkrBmx2B3BPRFwI3A58LnnuHOA24C3AKuA2SbPTqrVS/fM/rHb/g5mNAWkeQawCdkREa0R0Ad8Brh6wzQrg4eTxIyXrfwN4MCI6I2If8CCwNsVaK9KcL/DGM2cwe+qErEsxM0tdmgGxANhVstyWtJXaBFyTPH4PMF3S3Aqfi6QbJbVIamlvbx+xwss51t1Ly859Pr1kZmNG1p3UNwFXSHoCuALYDfRW+uSIuDMiVkbEyvnz56dVIwBPPL+frp4+B4SZjRmNKb72bmBhyXJT0vaKiNhDcgQhaRrw3ojYL2k3cOWA5z6aYq1DyuU7GCe4bOmcLMswMxs1aR5BbATOkbRU0gTgOuC+0g0kzZPUX8MtwF3J4weAd0qanXROvzNpy0xzvsAFTbOYMWl8lmWYmY2a1AIiInqAj1H8xf40cG9EbJV0u6Srks2uBLZLegY4Hfhs8txO4DMUQ2YjcHvSlokjXT08uWu/Ty+Z2ZiS5ikmImI9sH5A260lj9cB607w3Lt49YgiUxuf20dPX3h4DTMbU7LupK4JzfkOxjeIlUsyvxXDzGzUOCAqkMsXuHjhbKZMSPWAy8ysqjgghnDgaDdP7T7AGvc/mNkY44AYws+f7aQvcECY2ZjjgBhCc76DiY3juHjRrKxLMTMbVQ6IIeTyBS5bMoeJjQ1Zl2JmNqocEIMoHDrOL1582aeXzGxMckAMYkNr8d48B4SZjUUOiEHkWjuYNrGRCxfMzLoUM7NR54AYRHO+wGVLZtPY4L8mMxt7/JvvBF46eIzW9sOef9rMxiwHxAn0Ty/q/gczG6scECfQnO9g5uTxrDhzRtalmJllwgFxAs35AquXzWHcOGVdiplZJhwQZezqPELbvqPufzCzMc0BUUZ//4MnCDKzscwBUUZzvoN50yZy9mnTsi7FzCwzDogBIoLmfIE1y+ciuf/BzMYuB8QArR2H2fvycZ9eMrMxzwExQHP//Q+ef9rMxjgHxAC5fAdnzZzE4rlTsi7FzCxTDogSfX3BhtZO1iyf5/4HMxvzHBAltr/0Mp2Huzy8hpkZKQeEpLWStkvaIenmMusXSXpE0hOSNkt6d9I+XtI3JG2R9LSkW9Kss1+zx18yM3tFagEhqQH4MvAuYAVwvaQVAzb7S+DeiLgYuA74StL+O8DEiLgAuBT4Q0lL0qq1Xy7fwZK5U1gwa3Lab2VmVvXSPIJYBeyIiNaI6AK+A1w9YJsA+kfDmwnsKWmfKqkRmAx0AQdTrJWe3j5+lvQ/mJlZugGxANhVstyWtJX6NPC7ktqA9cDHk/Z1wGHgBeB54I6I6Bz4BpJulNQiqaW9vf2Uit265yAvH+/x6SUzs0TWndTXA3dHRBPwbuCfJI2jePTRC5wFLAU+JWnZwCdHxJ0RsTIiVs6fP/+UCvH9D2Zmr5VmQOwGFpYsNyVtpX4fuBcgInLAJGAe8H7g/ojojoi9wP8DVqZYK7nWAueePo350yem+TZmZjUjzYDYCJwjaamkCRQ7oe8bsM3zwDsAJL2RYkC0J+1vT9qnAquBX6RVaFdPHxuf7fTRg5lZidQCIiJ6gI8BDwBPU7xaaauk2yVdlWz2KeAPJG0Cvg3cEBFB8eqnaZK2Ugyar0fE5rRq3dS2n6Pdve6gNjMr0Zjmi0fEeoqdz6Vtt5Y83ga8tczzDlG81HVU5PIFJFi9bM5ovaWZWdXLupO6KjTnO1hx5gxmTZmQdSlmZlVjzAfEse5eHt+538N7m5kNMOYD4uCxbtaefwa/dt5pWZdiZlZVUu2DqAWnTZ/EF6+/OOsyzMyqzpg/gjAzs/IcEGZmVpYDwszMynJAmJlZWQ4IMzMrywFhZmZlOSDMzKwsB4SZmZWl4uCptU9SO7DzFF5iHtAxQuVkqV72A7wv1ape9qVe9gNObV8WR0TZGdfqJiBOlaSWiEh1UqLRUC/7Ad6XalUv+1Iv+wHp7YtPMZmZWVkOCDMzK8sB8ao7sy5ghNTLfoD3pVrVy77Uy35ASvviPggzMyvLRxBmZlaWA8LMzMoacwEh6S5JeyU9VdL2qKSau9xN0kJJj0jaJmmrpE8k7TW3P5ImSfq5pE3Jvvy3pP05SfOyru9kSJolaZ2kX0h6WtKaWvk3OcHn46+Tfdks6fuSZiXtN0j6UmbFDuFEn4+S9Z+SFP3/vyR9WtJN2VQ7NEkNkp6Q9K/J8jskPS7pSUn/V9LZSfvdkn57JN5zzAUEcDewNusiRkgP8KmIWAGsBj4qaUXGNQ3XceDtEfFm4CJgraTV2ZY0bP8LuD8izgPeDDydcT0n425e//l4EDg/Ii4EngFuGe2ihumEnw9JC4F3As9nWN/J+gSv/b/098AHIuIi4H8DfznSbzjmAiIifgJ0lln1O8k32GckvW206xqOiHghIh5PHr9M8T/PgmR1Te1PFB1KFscnP/1XUHw8+aa0RdJ52VRYGUkzgV8F/hEgIroiYn+yuur/Tcp9PiLiRxHRkyxuAJpKVp8l6X5Jv5T0hdGqsxJDfD7+J/BnvPp/rN+K5GivVdKfjF61g5PUBPwm8LWS5gBmJI9nAntK1v2qpOZkP4Z9NDHmAmIQjRGxCvgkcFvGtZw0SUuAi4GfJU01tz/JIfSTwF7gwYjo35eOiLiE4jemqj0FkFgKtANfT04HfE3S1GRdzf2blPER4N9Kli8CrgUuAK5NvplXndLPh6Srgd0RsanMpucBvwGsAm6TNH70qhzU31IMtL6Stv8MrJfUBnwQ+O8l684EfgX4rQHtJ8UB8arvJX8+BizJsI6TJmka8F3gkxFxMGmuuf2JiN7kcLkJWCXp/GRVLe1LI3AJ8PcRcTFwGLg5WVdL+/E6kv6C4mmbb5U0PxQRByLiGLANWJxJcYMo/XxQrP+/AreeYPMfRsTxiOig+EXl9FEpchCSfgvYGxGPDVj1p8C7I6IJ+DrwNyXr/iUi+iJiG6ewD43DfWIdOp782UsN/b0k33C+C3wrIr5Xsqom9wcgIvZLeoRXz4XX0r60AW0lRz/reDUgamk/XkPSDRS/jb4jXnvz1PGSx1W3XwM/H5IuoHiUt0kSFL+MPC5pVfKUatyftwJXSXo3MAmYIemHwHkl/8/+Gbi/5Dml+6HhvrGPIGqYiv/D/xF4OiL+Zqjtq5mk+SVXx0wGfh34RaZFDUNEvAjskvSGpOkdFL9Z1yxJayme3rgqIo5kXU+lyn0+ImJLRJwWEUsiYgnFQL8k+XerShFxS0Q0JfVeBzwMXA3MlHRustmvk8LFENWQjqNK0reBK4F5ybm7Wj0XDMVvFh8EtiTn7qF4+FyLzgS+IamB4heXeyPiX6v5MspBfBz4lqQJQCvwYeD72ZZUmRN8Pm4BJgIPJt+6N0TEH2VWZOXKfj4iYn12JY2MiOiR9AfAdyX1Afso9g+NKA+1YWZmZfkUk5mZleWAMDOzshwQZmZWlgPCzMzKckCYmVlZDgirS5LmJqNcPinpRUm7k8eHJH0lhff7dMl7/FLS90oHTkyG3DjhQIrJyKhnjXRdZqfCl7la3ZP0aeBQRNwxWu8h6VqKo7peEBHtFTz/UeCmiGhJq0azk+UjCBtTJF1ZMp7+pyV9Q9JPJe2UdI2kLySjxt7fP1CbpEsl/VjSY5IekHTmUO8TEf8M/Ah4f/Iaj0pamQxIeLekp5L3+dNktM2VFG+ue1LSZEm3StqYbHdncldw/+t8fuCosMnr3pFsv1nSx4dbu1k/B4SNdcuBtwNXAd8EHomIC4CjwG8mIfF3wG9HxKXAXcBnK3ztxymODlrqImBBRJyfvM/XI2Id0EIytn9EHAW+FBGXRcT5wGSK4yD1Kzcq7I0UBwC8KJm34VunWLvZ2Btqw2yAf4uIbklbgAZeHfBsC8VfuG8AzufVYSYagBcqfO1yg6S1Assk/R3wQ4pHGeX8mqQ/A6YAc4CtwA+SdeVGhf0PwFf7522IiM5kNNzh1m7mgLAx7zhARPRJ6i4ZqbSP4udDwNaIWDOM176Y4pHBKyJin6Q3U5xz4I+A9zFgDB1Jk4CvACsjYlfSvzFpYM0MPdroqdRu5lNMZkPYDsyXtAaKw0dLetNQT5L0XopTWn57QPs8YFxEfJfiFJGXJKteBqYnj/vDoCOZy6CSGcEeBP5QUmPyPnOGW7tZPx9BmA0iIrqSTuQvqjidaCPF2b22ltn8TyX9LjAVeIriHNsDr2BaQHG2uf4vZ/3zO98NfFXSUWAN8A/Ja7wIbKyg1K8B5wKbJXUD/xARXzqJ2s1ex5e5mplZWT7FZGZmZTkgzMysLAeEmZmV5YAwM7OyHBBmZlaWA8LMzMpyQJiZWVn/H96/VNGxQw2tAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot\n",
    "\n",
    "# btc\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"recall\"] for key, value in performance.items() if \"btc\" in key],\n",
    "            label=\"btc\"\n",
    "           )\n",
    "# eth\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"recall\"] for key, value in performance.items() if \"eth\" in key],\n",
    "            label=\"eth\"\n",
    "           )\n",
    "# bnb\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"recall\"] for key, value in performance.items() if \"bnb\" in key],\n",
    "            label=\"bnb\"\n",
    "           )\n",
    "\n",
    "pyplot.xlabel(\"Time Distance\")\n",
    "pyplot.ylabel(\"recall\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "id": "6513a9f4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'precision')"
      ]
     },
     "execution_count": 96,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from matplotlib import pyplot\n",
    "\n",
    "# btc\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"f1-score\"] for key, value in performance.items() if \"btc\" in key],\n",
    "            label=\"btc\"\n",
    "           )\n",
    "# eth\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"f1-score\"] for key, value in performance.items() if \"eth\" in key],\n",
    "            label=\"eth\"\n",
    "           )\n",
    "# bnb\n",
    "pyplot.plot([\"1h\", \"2h\", \"3h\", \"6h\", \"12h\", \"24h\", \"48h\"],\n",
    "            [value[\"f1-score\"] for key, value in performance.items() if \"bnb\" in key],\n",
    "            label=\"bnb\"\n",
    "           )\n",
    "\n",
    "pyplot.xlabel(\"Time Distance\")\n",
    "pyplot.ylabel(\"f1-score\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "id": "28265b50",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.5,\n",
       " 0.5714285714285714,\n",
       " 0.5714285714285714,\n",
       " 0.5714285714285714,\n",
       " 0.7142857142857143,\n",
       " 0.5714285714285714,\n",
       " 0.4]"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[value[\"preciosion\"] for key, value in performance.items() if \"btc\" in key]"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}