Esempio n. 1
0
def main():
    # get api information
    api_file_path = 'api_key'
    api_key, api_secret = get_api_information(api_file_path)

    # initialize client
    client = Client(api_key, api_secret)

    # define symbol
    symbol = 'BTCUSDT'

    output_dir = f'../output/dataframes/{symbol}'

    # iterate over all intervals
    for interval in kline_intervals().keys():
        start_time = timer()

        # get data
        df = get_all_candles(client, symbol, interval,
                             client._get_earliest_valid_timestamp(symbol, interval))

        # store data
        df.to_csv(f'{output_dir}/{symbol}_{interval[-1]}_{interval[:-1]}.csv', index=False)
        print(f'Done with interval {interval} after {timer() - start_time}s.')
Esempio n. 2
0
    ''' define how to process incoming WebSocket messages '''
    if msg['e'] != 'error':
        print(msg['c'])
        btc_price['last'] = msg['c']
        btc_price['bid'] = msg['b']
        btc_price['last'] = msg['a']
    else:
        btc_price['error'] = True


bsm = BinanceSocketManager(client)
conn_key = bsm.start_symbol_ticker_socket('BTCUSDT', btc_trade_history)

bsm.start()

timestamp = client._get_earliest_valid_timestamp('BTCUSDT', '1m')
print(timestamp)

# bsm.stop_socket(conn_key)

# request historical candle (or klines) data
# bars = client.get_historical_klines('BTCUSDT','1m', timestamp, limit=1000)

bars = client.get_historical_klines("BTCUSDT", Client.KLINE_INTERVAL_1MINUTE,
                                    "1 day ago UTC")

with open('btc_bars.csv', 'w', newline='') as f:
    wr = csv.writer(f)
    for line in bars:
        wr.writerow(line)
Esempio n. 3
0
class Data:
    def __init__(self,
                 interval: str = '1h',
                 symbol: str = 'BTCUSDT',
                 loadData: bool = True,
                 updateData: bool = True,
                 log: bool = False,
                 logFile: str = 'data',
                 logObject=None,
                 precision: int = 2,
                 callback=None,
                 caller=None):
        """
        Data object that will retrieve current and historical prices from the Binance API and calculate moving averages.
        :param: interval: Interval for which the data object will track prices.
        :param: symbol: Symbol for which the data object will track prices.
        :param: loadData: Boolean for whether data will be loaded or not.
        :param: updateData: Boolean for whether data will be updated if it is loaded.
        :param: precision: Precision to round data to.
        :param: callback: Signal for GUI to emit back to (if passed).
        :param: caller: Caller of callback (if passed).
        """
        self.callback = callback  # Used to emit signals to GUI if provided.
        self.caller = caller  # Used to specify which caller emitted signals for GUI.
        self.binanceClient = Client(
        )  # Initialize Binance client to retrieve data.
        self.logger = self.get_logging_object(enable_logging=log,
                                              logFile=logFile,
                                              loggerObject=logObject)
        self.validate_interval(interval)  # Validate the interval provided.
        self.interval = interval  # Interval to trade in.
        self.intervalUnit, self.intervalMeasurement = self.get_interval_unit_and_measurement(
        )
        self.precision = precision  # Decimal precision with which to show data.
        self.dataLimit = 2000  # Max amount of data to contain.

        self.downloadCompleted = False  # Boolean to determine whether data download is completed or not.
        self.downloadLoop = True  # Boolean to determine whether data is being downloaded or not.

        self.tickers = self.binanceClient.get_all_tickers(
        )  # A list of all the tickers on Binance.
        self.symbol = symbol.upper()  # Symbol of data being used.
        self.validate_symbol(self.symbol)  # Validate symbol.
        self.data = []  # Total bot data.
        self.ema_dict = {}  # Cached past EMA data for memoization.
        self.rsi_data = {}  # Cached past RSI data for memoization.
        self.current_values = {  # This dictionary will hold current data values.
            'date_utc': datetime.now(tz=timezone.utc),
            'open': 0,
            'high': 0,
            'low': 0,
            'close': 0,
            'volume': 0,
            'quote_asset_volume': 0,
            'number_of_trades': 0,
            'taker_buy_base_asset': 0,
            'taker_buy_quote_asset': 0
        }

        self.databaseTable = f'data_{self.interval}'
        self.databaseFile = self.get_database_file()
        self.create_table()

        if loadData:
            # Create, initialize, store, and get values from database.
            self.load_data(update=updateData)

    @staticmethod
    def get_logging_object(enable_logging: bool, logFile: str, loggerObject):
        """
        Returns a logger object.
        :param enable_logging: Boolean that determines whether logging is enabled or not.
        :param logFile: File to log to.
        :param loggerObject: Logger object to return if there is one already specified.
        :return: Logger object or None.
        """
        if loggerObject:
            return loggerObject

        if enable_logging:
            return get_logger(log_file=logFile, logger_name=logFile)

        return None

    def validate_interval(self, interval: str):
        """
        Validates interval. If incorrect interval, raises ValueError.
        :param interval: Interval to be checked.
        """
        if not self.is_valid_interval(interval):
            raise ValueError(f'Invalid interval {interval} specified.')

    def validate_symbol(self, symbol: str):
        """
        Validates symbol for data to be retrieved. Raises ValueError if symbol type is incorrect.
        :param symbol: Symbol to be checked.
        """
        if not self.is_valid_symbol(symbol):
            raise ValueError(f'Invalid symbol {symbol} specified.')

    def load_data(self, update: bool = True):
        """
        Loads data to Data object.
        :param update: Boolean that determines whether data is updated or not.
        """
        self.get_data_from_database()
        if update:
            if not self.database_is_updated():
                self.output_message("Updating data...")
                self.update_database_and_data()
            else:
                self.output_message("Database is up-to-date.")

    def output_message(self,
                       message: str,
                       level=2,
                       printMessage: bool = False):
        """
        I need to research the logging module better, but in essence, this function just logs and optionally prints
        message provided.
        :param message: Messaged to be logged and potentially printed.
        :param level: Level message will be logged at.
        :param printMessage: Boolean that decides whether message will also be printed or not.
        """
        if printMessage:
            print(message)

        if self.logger:
            if level == 2:
                self.logger.info(message)
            elif level == 3:
                self.logger.debug(message)
            elif level == 4:
                self.logger.warning(message)
            elif level == 5:
                self.logger.critical(message)

    def get_database_file(self) -> str:
        """
        Retrieves database file path.
        :return: Database file path.
        """
        database_folder = os.path.join(ROOT_DIR, 'Databases')
        if not os.path.exists(database_folder):
            os.mkdir(database_folder)

        filePath = os.path.join(database_folder, f'{self.symbol}.db')
        return filePath

    def create_table(self):
        """
        Creates a new table with interval if it does not exist.
        """
        with closing(sqlite3.connect(self.databaseFile)) as connection:
            with closing(connection.cursor()) as cursor:
                cursor.execute(f'''
                                CREATE TABLE IF NOT EXISTS {self.databaseTable}(
                                date_utc TEXT PRIMARY KEY,
                                open_price TEXT NOT NULL,
                                high_price TEXT NOT NULL,
                                low_price TEXT NOT NULL,
                                close_price TEXT NOT NULL,
                                volume TEXT NOT NULL,
                                quote_asset_volume TEXT NOT NULL,
                                number_of_trades TEXT NOT NULL,
                                taker_buy_base_asset TEXT NOT NULL,
                                taker_buy_quote_asset TEXT NOT NULL
                                );''')
                connection.commit()

    def dump_to_table(self, totalData: list = None) -> bool:
        """
        Dumps date and price information to database.
        :return: A boolean whether data entry was successful or not.
        """
        if totalData is None:
            totalData = self.data

        query = f'''INSERT INTO {self.databaseTable} (date_utc, open_price, high_price, low_price, close_price,
                    volume, quote_asset_volume, number_of_trades, taker_buy_base_asset, taker_buy_quote_asset)
                    VALUES (?, ?, ?, ?, ?, ?, ?, ?, ?, ?);'''
        with closing(sqlite3.connect(self.databaseFile)) as connection:
            with closing(connection.cursor()) as cursor:
                for data in totalData:
                    try:
                        cursor.execute(query, (
                            data['date_utc'].strftime('%Y-%m-%d %H:%M:%S'),
                            data['open'],
                            data['high'],
                            data['low'],
                            data['close'],
                            data['volume'],
                            data['quote_asset_volume'],
                            data['number_of_trades'],
                            data['taker_buy_base_asset'],
                            data['taker_buy_quote_asset'],
                        ))
                    except sqlite3.IntegrityError:
                        pass  # This just means the data already exists in the database, so ignore.
                    except sqlite3.OperationalError:
                        connection.commit()
                        self.output_message(
                            "Insertion to database failed. Will retry next run.",
                            4)
                        return False
            connection.commit()
        self.output_message("Successfully stored all new data to database.")
        return True

    def get_latest_database_row(self) -> list:
        """
        Returns the latest row from database table.
        :return: Row data or None depending on if value exists.
        """
        with closing(sqlite3.connect(self.databaseFile)) as connection:
            with closing(connection.cursor()) as cursor:
                cursor.execute(
                    f'SELECT date_utc FROM {self.databaseTable} ORDER BY date_utc DESC LIMIT 1'
                )
                return cursor.fetchone()

    def get_data_from_database(self):
        """
        Loads data from database and appends it to run-time data.
        """
        with closing(sqlite3.connect(self.databaseFile)) as connection:
            with closing(connection.cursor()) as cursor:
                rows = cursor.execute(f'''
                        SELECT "date_utc", "open_price", "high_price", "low_price", "close_price", "volume",
                        "quote_asset_volume", "number_of_trades", "taker_buy_base_asset", "taker_buy_quote_asset"
                        FROM {self.databaseTable} ORDER BY date_utc DESC
                        ''').fetchall()

        if len(rows) > 0:
            self.output_message("Retrieving data from database...")
        else:
            self.output_message("No data found in database.")
            return

        for row in rows:
            date_utc = datetime.strptime(
                row[0], '%Y-%m-%d %H:%M:%S').replace(tzinfo=timezone.utc)
            self.data.append({
                'date_utc': date_utc,
                'open': float(row[1]),
                'high': float(row[2]),
                'low': float(row[3]),
                'close': float(row[4]),
                'volume': float(row[5]),
                'quote_asset_volume': float(row[6]),
                'number_of_trades': float(row[7]),
                'taker_buy_base_asset': float(row[8]),
                'taker_buy_quote_asset': float(row[9]),
            })

    def database_is_updated(self) -> bool:
        """
        Checks if data is updated or not with database by interval provided in accordance to UTC time.
        :return: A boolean whether data is updated or not.
        """
        result = self.get_latest_database_row()
        if result is None:
            return False
        latestDate = datetime.strptime(
            result[0], '%Y-%m-%d %H:%M:%S').replace(tzinfo=timezone.utc)
        return self.is_latest_date(latestDate)

    # noinspection PyProtectedMember
    def get_latest_timestamp(self) -> int:
        """
        Returns latest timestamp available based on database.
        :return: Latest timestamp.
        """
        result = self.get_latest_database_row()
        if result is None:
            return self.binanceClient._get_earliest_valid_timestamp(
                self.symbol, self.interval)
        else:
            latestDate = datetime.strptime(
                result[0], '%Y-%m-%d %H:%M:%S').replace(tzinfo=timezone.utc)
            return int(latestDate.timestamp()
                       ) * 1000 + 1  # Converting timestamp to milliseconds

    # noinspection PyProtectedMember
    def update_database_and_data(self):
        """
        Updates database by retrieving information from Binance API
        """
        result = self.get_latest_database_row()
        if result is None:  # Then get the earliest timestamp possible
            timestamp = self.binanceClient._get_earliest_valid_timestamp(
                self.symbol, self.interval)
            self.output_message(
                f'Downloading all available historical data for {self.interval} intervals.'
            )
        else:
            latestDate = datetime.strptime(
                result[0], '%Y-%m-%d %H:%M:%S').replace(tzinfo=timezone.utc)
            timestamp = int(latestDate.timestamp()
                            ) * 1000  # Converting timestamp to milliseconds
            dateWithIntervalAdded = latestDate + timedelta(
                minutes=self.get_interval_minutes())
            self.output_message(
                f"Previous data up to UTC {dateWithIntervalAdded} found.")

        if not self.database_is_updated():
            newData = self.get_new_data(timestamp)
            self.output_message("Successfully downloaded all new data.")
            self.output_message("Inserting data to live program...")
            self.insert_data(newData)
            self.output_message("Storing updated data to database...")
            self.dump_to_table(self.data[-len(newData):])
        else:
            self.output_message("Database is up-to-date.")

    # noinspection PyProtectedMember
    def custom_get_new_data(self,
                            limit: int = 500,
                            progress_callback=None,
                            locked=None,
                            removeFirst=False,
                            caller=-1) -> List[dict]:
        """
        Returns new data from Binance API from timestamp specified, however this one is custom-made.
        :param caller: Caller that called this function. Only used for botThread.
        :param removeFirst: Boolean whether newest data is removed or not.
        :param locked: Signal to emit back to GUI when storing data. Cannot be canceled once here. Used for databases.
        :param progress_callback: Signal to emit back to GUI to show progress.
        :param limit: Limit per pull.
        :return: A list of dictionaries.
        """
        # This code below is taken from binance client and slightly refactored.
        self.downloadLoop = True
        output_data = []  # Initialize our list
        timeframe = interval_to_milliseconds(self.interval)
        start_ts = total_beginning_timestamp = self.get_latest_timestamp()
        end_progress = time.time() * 1000 - total_beginning_timestamp
        idx = 0

        while True and self.downloadLoop:
            tempData = self.binanceClient.get_klines(symbol=self.symbol,
                                                     interval=self.interval,
                                                     limit=limit,
                                                     startTime=start_ts,
                                                     endTime=None)

            if not len(tempData):
                break

            output_data += tempData
            start_ts = tempData[-1][0]
            if progress_callback:
                progress = (start_ts -
                            total_beginning_timestamp) / end_progress * 94
                progress_callback.emit(int(progress), "Downloading data...",
                                       caller)

            idx += 1
            # check if we received less than the required limit and exit the loop
            if len(tempData) < limit:
                # exit the while loop
                break

            # increment next call by our timeframe
            start_ts += timeframe

            # sleep after every 5th call to be kind to the API
            if idx % 5 == 0:
                time.sleep(1)

        if not self.downloadLoop:
            progress_callback.emit(-1, "Download canceled.", caller)
            return []

        if locked:
            locked.emit()

        if removeFirst:  # This should be refactored once data is inserted in the reverse order.
            output_data.pop()

        progress_callback.emit(95, "Saving data...", caller)
        self.insert_data(output_data)
        progress_callback.emit(
            97, "This may take a while. Dumping data to database...", caller)

        if removeFirst:  # We don't want current data as it's not the latest data.
            self.dump_to_table(self.data[:len(output_data)])
        else:
            self.dump_to_table(self.data[1:len(output_data)])

        progress_callback.emit(100, "Downloaded all new data successfully.",
                               caller)
        self.downloadLoop = False
        self.downloadCompleted = True
        return self.data

    def get_new_data(self, timestamp: int, limit: int = 1000) -> list:
        """
        Returns new data from Binance API from timestamp specified.
        :param timestamp: Initial timestamp.
        :param limit: Limit per pull.
        :return: A list of dictionaries.
        """
        newData = self.binanceClient.get_historical_klines(self.symbol,
                                                           self.interval,
                                                           timestamp + 1,
                                                           limit=limit)
        self.downloadCompleted = True
        return newData[:
                       -1]  # Up to -1st index, because we don't want current period data.

    def is_latest_date(self, latestDate: datetime) -> bool:
        """
        Checks whether the latest date available is the latest period available.
        :param latestDate: Datetime object.
        :return: True or false whether date is latest period or not.
        """
        minutes = self.get_interval_minutes()
        current_date = latestDate + timedelta(minutes=minutes) + timedelta(
            seconds=5)  # 5s leeway for server update
        return current_date >= datetime.now(
            timezone.utc) - timedelta(minutes=minutes)

    def data_is_updated(self) -> bool:
        """
        Checks whether data is fully updated or not.
        :return: A boolean whether data is updated or not with Binance values.
        """
        latestDate = self.data[0]['date_utc']
        return self.is_latest_date(latestDate)

    def insert_data(self, newData: list):
        """
        Inserts data from newData to run-time data.
        :param newData: List with new data values.
        """
        temp_data = []

        for data in newData[::-1]:
            parsedDate = datetime.fromtimestamp(int(data[0]) / 1000,
                                                tz=timezone.utc)
            current_dict = {
                'date_utc': parsedDate,
                'open': float(data[1]),
                'high': float(data[2]),
                'low': float(data[3]),
                'close': float(data[4]),
                'volume': float(data[5]),
                'quote_asset_volume': float(data[6]),
                'number_of_trades': float(data[7]),
                'taker_buy_base_asset': float(data[8]),
                'taker_buy_quote_asset': float(data[9]),
            }
            temp_data.append(current_dict)

        self.data = temp_data + self.data

    def update_data(self, verbose: bool = False):
        """
        Updates run-time data with Binance API values.
        """
        latestDate = self.data[0]['date_utc']
        timestamp = int(latestDate.timestamp()) * 1000
        dateWithIntervalAdded = latestDate + timedelta(
            minutes=self.get_interval_minutes())
        if verbose:
            self.output_message(
                f"Previous data found up to UTC {dateWithIntervalAdded}.")
        if not self.data_is_updated():
            # self.try_callback("Found new data. Attempting to update...")
            newData = []
            while len(newData) == 0:
                time.sleep(
                    0.5
                )  # Sleep half a second for server to refresh new values.
                newData = self.get_new_data(timestamp)
            self.insert_data(newData)
            if verbose:
                self.output_message("Data has been updated successfully.\n")
            # self.try_callback("Updated data successfully.")
        else:
            self.output_message("Data is up-to-date.\n")

    def remove_past_data_if_needed(self):
        """
        Remove past data past data limit.
        """
        if len(self.data) > self.dataLimit:  # Remove past data.
            self.dump_to_table()
            self.data = self.data[:self.dataLimit // 2]

    def get_current_data(self, counter: int = 0) -> dict:
        """
        Retrieves current market dictionary with open, high, low, close prices.
        :param counter: Counter to check how many times bot is trying to retrieve current data.
        :return: A dictionary with current open, high, low, and close prices.
        """
        try:
            self.remove_past_data_if_needed()
            if not self.data_is_updated():
                self.update_data()

            currentInterval = self.data[0]['date_utc'] + timedelta(
                minutes=self.get_interval_minutes())
            currentTimestamp = int(currentInterval.timestamp() * 1000)

            nextInterval = currentInterval + timedelta(
                minutes=self.get_interval_minutes())
            nextTimestamp = int(nextInterval.timestamp() * 1000) - 1
            currentData = self.binanceClient.get_klines(
                symbol=self.symbol,
                interval=self.interval,
                startTime=currentTimestamp,
                endTime=nextTimestamp,
            )[0]
            currentDataDictionary = {
                'date_utc': currentInterval,
                'open': float(currentData[1]),
                'high': float(currentData[2]),
                'low': float(currentData[3]),
                'close': float(currentData[4]),
                'volume': float(currentData[5]),
                'quote_asset_volume': float(currentData[6]),
                'number_of_trades': float(currentData[7]),
                'taker_buy_base_asset': float(currentData[8]),
                'taker_buy_quote_asset': float(currentData[9]),
            }
            self.current_values = currentDataDictionary
            if counter > 0:
                self.try_callback("Successfully reconnected.")
            return currentDataDictionary
        except Exception as e:
            sleepTime = 5 + counter * 2
            error_message = f"Error: {e}. Retrying in {sleepTime} seconds..."
            self.output_message(error_message, 4)
            self.try_callback(
                f"Internet connectivity issue detected. Trying again in {sleepTime} seconds."
            )
            self.ema_dict = {}  # Reset EMA cache as it could be corrupted.
            time.sleep(sleepTime)
            return self.get_current_data(counter=counter + 1)

    def try_callback(self, message: str):
        """
        Attempts to emit a signal to the GUI that called this data object (if it was called by a GUI).
        :param message: Message to send back.
        """
        if self.callback and self.caller is not None:
            self.callback.emit(self.caller, message)

    def get_current_price(self) -> float:
        """
        Returns the current market ticker price.
        :return: Ticker market price
        """
        try:
            return float(
                self.binanceClient.get_symbol_ticker(
                    symbol=self.symbol)['price'])
        except Exception as e:
            error_message = f'Error: {e}. Retrying in 15 seconds...'
            self.output_message(error_message, 4)
            self.try_callback(message=error_message)
            time.sleep(15)
            return self.get_current_price()

    def get_interval_unit_and_measurement(self) -> Tuple[str, int]:
        """
        Returns interval unit and measurement.
        :return: A tuple with interval unit and measurement respectively.
        """
        unit = self.interval[-1]  # Gets the unit of the interval. eg 12h = h
        measurement = int(
            self.interval[:-1])  # Gets the measurement, eg 12h = 12
        return unit, measurement

    def get_interval_minutes(self) -> int:
        """
        Returns interval minutes.
        :return: An integer representing the minutes for an interval.
        """
        if self.intervalUnit == 'h':
            return self.intervalMeasurement * 60
        elif self.intervalUnit == 'm':
            return self.intervalMeasurement
        elif self.intervalUnit == 'd':
            return self.intervalMeasurement * 24 * 60
        else:
            raise ValueError("Invalid interval.", 4)

    def create_folders_and_change_path(self, folderName: str):
        """
        Creates appropriate folders for data storage then changes current working directory to it.
        :param folderName: Folder to create.
        """
        os.chdir(ROOT_DIR)
        if not os.path.exists(
                folderName):  # Create CSV folder if it doesn't exist
            os.mkdir(folderName)
        os.chdir(folderName)  # Go inside the folder.

        if not os.path.exists(
                self.symbol
        ):  # Create symbol folder inside CSV folder if it doesn't exist.
            os.mkdir(self.symbol)
        os.chdir(self.symbol)  # Go inside the folder.

    def write_csv_data(self,
                       totalData: list,
                       fileName: str,
                       armyTime: bool = True) -> str:
        """
        Writes CSV data to CSV folder in root directory of application.
        :param armyTime: Boolean if date will be in army type. If false, data will be in standard type.
        :param totalData: Data to write to CSV file.
        :param fileName: Filename to name CSV in.
        :return: Absolute path to CSV file.
        """
        currentPath = os.getcwd()
        self.create_folders_and_change_path(folderName="CSV")

        with open(fileName, 'w') as f:
            f.write(
                "Date_UTC, Open, High, Low, Close, Volume, Quote_Asset_Volume, Number_of_Trades, "
                "Taker_Buy_Base_Asset, Taker_Buy_Quote_Asset\n")
            for data in totalData:
                if armyTime:
                    parsedDate = data['date_utc'].strftime("%m/%d/%Y %H:%M")
                else:
                    parsedDate = data['date_utc'].strftime("%m/%d/%Y %I:%M %p")
                f.write(
                    f'{parsedDate}, {data["open"]}, {data["high"]}, {data["low"]}, {data["close"]}, '
                    f'{data["volume"]}, {data["quote_asset_volume"]}, {data["number_of_trades"]}, '
                    f'{data["taker_buy_base_asset"]}, {data["taker_buy_quote_asset"]}\n'
                )

        path = os.path.join(os.getcwd(), fileName)
        os.chdir(currentPath)

        return path

    def create_csv_file(self,
                        descending: bool = True,
                        armyTime: bool = True,
                        startDate: datetime = None) -> str:
        """
        Creates a new CSV file with current interval and returns the absolute path to file.
        :param startDate: Date to have CSV data from.
        :param descending: Boolean that decides whether values in CSV are in descending format or not.
        :param armyTime: Boolean that dictates whether dates will be written in army-time format or not.
        """
        self.update_database_and_data()  # Update data if updates exist.
        fileName = f'{self.symbol}_data_{self.interval}.csv'

        data = self.data
        if startDate is not None:
            for index, period in enumerate(data):
                if period['date_utc'].date() < startDate:
                    data = self.data[:index]
                    break

        if descending:
            path = self.write_csv_data(data,
                                       fileName=fileName,
                                       armyTime=armyTime)
        else:
            path = self.write_csv_data(data[::-1],
                                       fileName=fileName,
                                       armyTime=armyTime)

        self.output_message(f'Data saved to {path}.')
        return path

    @staticmethod
    def get_custom_csv_data(symbol: str,
                            interval: str,
                            descending: bool = True) -> str:
        """
        Creates a new CSV file with interval specified and returns the absolute path of CSV file.
        :param symbol: Symbol to get data for.
        :param interval: Interval to get data for.
        :param descending: Returns data in specified sort. If descending, writes data from most recent to oldest data.
        """
        tempData = Data(interval=interval, symbol=symbol)
        return tempData.create_csv_file(descending=descending)

    def is_valid_interval(self, interval: str) -> bool:
        """
        Returns whether interval provided is valid or not.
        :param interval: Interval argument.
        :return: A boolean whether the interval is valid or not.
        """
        availableIntervals = ('12h', '15m', '1d', '1h', '1m', '2h', '30m',
                              '3d', '3m', '4h', '5m', '6h', '8h')
        if interval in availableIntervals:
            return True
        else:
            self.output_message(
                f'Invalid interval. Available intervals are: \n{availableIntervals}'
            )
            return False

    def is_valid_symbol(self, symbol: str) -> bool:
        """
        Checks whether the symbol provided is valid or not for Binance.
        :param symbol: Symbol to be checked.
        :return: A boolean whether the symbol is valid or not.
        """
        for ticker in self.tickers:
            if ticker['symbol'] == symbol:
                return True
        return False

    def is_valid_average_input(self,
                               shift: int,
                               prices: int,
                               extraShift: int = 0) -> bool:
        """
        Checks whether shift, prices, and (optional) extraShift are valid.
        :param shift: Periods from current period.
        :param prices: Amount of prices to iterate over.
        :param extraShift: Extra shift for EMA.
        :return: A boolean whether shift, prices, and extraShift are logical or not.
        """
        if shift < 0:
            self.output_message("Shift cannot be less than 0.")
            return False
        elif prices <= 0:
            self.output_message("Prices cannot be 0 or less than 0.")
            return False
        elif shift + extraShift + prices > len(self.data) + 1:
            self.output_message(
                "Shift + prices period cannot be more than data available.")
            return False
        return True

    def verify_integrity(self) -> bool:
        """
        Verifies integrity of data by checking if there's any repeated data.
        :return: A boolean whether the data contains no repeated data or not.
        """
        if len(self.data) < 1:
            self.output_message("No data found.", 4)
            return False

        previousData = self.data[0]
        for data in self.data[1:]:
            if data['date_utc'] == previousData['date_utc']:
                self.output_message("Repeated data detected.", 4)
                self.output_message(f'Previous data: {previousData}', 4)
                self.output_message(f'Next data: {data}', 4)
                return False
            previousData = data

        self.output_message("Data has been verified to be correct.")
        return True

    def get_total_non_updated_data(self) -> DATA_TYPE:
        return [self.current_values] + self.data

    def get_summation(self,
                      prices: int,
                      parameter: str,
                      round_value: bool = True,
                      update: bool = True) -> float:
        """
        Returns total summation.
        :param update: Boolean for whether function should call API and get latest data or not.
        :param prices: Amount of periods to iterate through for summation.
        :param parameter: Parameter to iterate through.
        :param round_value: Boolean that determines whether returned output is rounded or not.
        :return: Total summation.
        """
        data = [self.get_current_data()
                ] + self.data if update else self.get_total_non_updated_data()
        data = data[:prices]

        total = 0
        for period in data:
            total += period[parameter]

        if round_value:
            return round(total, self.precision)
        return total

    def get_lowest_low_value(self,
                             prices: int,
                             parameter: str = 'low',
                             round_value: bool = True,
                             update: bool = True) -> float:
        """
        Function that returns the lowest low values.
        :param update: Boolean for whether function should call API and get latest data or not.
        :param prices: Amount of periods to iterate through.
        :param parameter: Parameter to iterate through. By default, it is low.
        :param round_value: Boolean that determines whether returned output is rounded or not.
        :return: Lowest low value from periods.
        """
        data = [self.get_current_data()
                ] + self.data if update else self.get_total_non_updated_data()
        data = data[:prices]

        lowest = data[0][parameter]

        for period in data[1:]:
            if period[parameter] < lowest:
                lowest = period[parameter]

        if round_value:
            return round(lowest, self.precision)
        return lowest

    def get_highest_high_value(self,
                               prices: int,
                               parameter: str = 'high',
                               round_value: bool = True,
                               update: bool = True) -> float:
        """
        Function that returns the highest high values.
        :param update: Boolean for whether function should call API and get latest data or not.
        :param prices: Amount of periods to iterate through.
        :param parameter: Parameter to iterate through. By default, it is high.
        :param round_value: Boolean that determines whether returned output is rounded or not.
        :return: Highest high value from periods.
        """
        data = [self.get_current_data()
                ] + self.data if update else self.get_total_non_updated_data()
        data = data[:prices]

        highest = data[0][parameter]

        for period in data[1:]:
            if period[parameter] > highest:
                highest = period[parameter]

        if round_value:
            return round(highest, self.precision)
        return highest

    @staticmethod
    def helper_get_ema(up_data: list, down_data: list, periods: int) -> tuple:
        """
        Helper function to get the EMA for relative strength index.
        :param down_data: Other data to get EMA of.
        :param up_data: Data to get EMA of.
        :param periods: Number of periods to iterate through.
        :return: EMA
        """
        emaUp = up_data[0]
        emaDown = down_data[0]
        alpha = 1 / periods

        for index in range(1, len(up_data)):
            emaUp = up_data[index] * alpha + emaUp * (1 - alpha)
            emaDown = down_data[index] * alpha + emaDown * (1 - alpha)

        return emaUp, emaDown

    def get_rsi(self,
                prices: int = 14,
                parameter: str = 'close',
                shift: int = 0,
                round_value: bool = True,
                update: bool = True) -> float:
        """
        Returns relative strength index.
        :param update: Boolean for whether function should call API and get latest data or not.
        :param prices: Amount of prices to iterate through.
        :param parameter: Parameter to use for iterations. By default, it's close.
        :param shift: Amount of prices to shift prices by.
        :param round_value: Boolean that determines whether final value is rounded or not.
        :return: Final relative strength index.
        """
        if not self.is_valid_average_input(shift, prices):
            raise ValueError('Invalid input specified.')

        if shift > 0:
            updateDict = False
            data = self.data
            shift -= 1
        else:
            updateDict = True
            data = [self.get_current_data(
            )] + self.data if update else self.get_total_non_updated_data()

        start = 500 + prices + shift if len(
            data) > 500 + prices + shift else len(data)
        data = data[shift:start]
        data = data[:]
        data.reverse()

        ups, downs = get_ups_and_downs(data=data, parameter=parameter)
        averageUp, averageDown = self.helper_get_ema(ups, downs, prices)
        rs = averageUp / averageDown
        rsi = 100 - 100 / (1 + rs)

        if shift == 0 and updateDict:
            self.rsi_data[prices] = rsi

        if round_value:
            return round(rsi, self.precision)
        return rsi

    def get_sma(self,
                prices: int,
                parameter: str,
                shift: int = 0,
                round_value: bool = True,
                update: bool = True) -> float:
        """
        Returns the simple moving average with run-time data and prices provided.
        :param update: Boolean for whether function should call API and get latest data or not.
        :param boolean round_value: Boolean that specifies whether return value should be rounded
        :param int prices: Number of values for average
        :param int shift: Prices shifted from current price
        :param str parameter: Parameter to get the average of (e.g. open, close, high or low values)
        :return: SMA
        """
        if not self.is_valid_average_input(shift, prices):
            raise ValueError('Invalid average input specified.')

        data = [self.get_current_data()
                ] + self.data if update else self.get_total_non_updated_data()
        data = data[
            shift:prices +
            shift]  # Data now starts from shift and goes up to prices + shift
        sma = get_sma(data, prices, parameter)

        if round_value:
            return round(sma, self.precision)
        return sma

    def get_wma(self,
                prices: int,
                parameter: str,
                shift: int = 0,
                round_value: bool = True,
                update: bool = True) -> float:
        """
        Returns the weighted moving average with run-time data and prices provided.
        :param update: Boolean for whether function should call API and get latest data or not.
        :param shift: Prices shifted from current period.
        :param boolean round_value: Boolean that specifies whether return value should be rounded
        :param int prices: Number of prices to loop over for average
        :param parameter: Parameter to get the average of (e.g. open, close, high or low values)
        :return: WMA
        """
        if not self.is_valid_average_input(shift, prices):
            raise ValueError('Invalid average input specified.')

        data = [self.get_current_data()
                ] + self.data if update else self.get_total_non_updated_data()
        data = data[shift:prices + shift]
        wma = get_wma(data, prices, parameter)

        if round_value:
            return round(wma, self.precision)
        return wma

    def get_ema(self,
                prices: int,
                parameter: str,
                shift: int = 0,
                sma_prices: int = 5,
                round_value: bool = True,
                update: bool = True) -> float:
        """
        Returns the exponential moving average with data provided.
        :param update: Boolean for whether function should call API and get latest data or not.
        :param shift: Prices shifted from current period.
        :param round_value: Boolean that specifies whether return value should be rounded
        :param int sma_prices: SMA prices to get first EMA over
        :param int prices: Days to iterate EMA over (or the period)
        :param str parameter: Parameter to get the average of (e.g. open, close, high, or low values)
        :return: EMA
        """
        if not self.is_valid_average_input(shift, prices, sma_prices):
            raise ValueError('Invalid average input specified.')
        elif sma_prices <= 0:
            raise ValueError(
                "Initial amount of SMA values for initial EMA must be greater than 0."
            )

        if not self.data_is_updated(
        ):  # Check if data is valid. If not, memoized data will be corrupted.
            self.ema_dict = {}
            self.update_data()

        data = [self.get_current_data()
                ] + self.data if update else self.get_total_non_updated_data()
        data = data[shift:]
        ema, self.ema_dict = get_ema(data, prices, parameter, sma_prices,
                                     self.ema_dict)

        if round_value:
            return round(ema, self.precision)
        return ema
import backtrader.indicators as btind

key = ""
secret = ""
client = Client(key, secret)
currencies = ['BTCUSDT', 'BCHUSDT', 'ETHUSDT', 'XRPUSDT', 'BNBUSDT']

bitcoin = client.get_historical_klines("BTCUSDT",
                                       Client.KLINE_INTERVAL_15MINUTE,
                                       "20 Oct, 2020")

df = {}
timestamp = []
timeperiod = '1h'
for crypto in currencies:
    timestamp.append(client._get_earliest_valid_timestamp(crypto, timeperiod))

timestamp = max(timestamp)
startdate = timestamp

for crypto in currencies:
    klines = client.get_historical_klines(crypto, timeperiod, startdate)

    for line in klines:
        del line[6:]

    df[crypto] = pd.DataFrame(
        klines, columns=['date', 'open', 'high', 'low', 'close', 'volume'])
    df[crypto]['date'] = pd.to_datetime(df[crypto]['date'], unit='ms')
    df[crypto].set_index('date', inplace=True)
Esempio n. 5
0
def BTC_alarm(request):
	#GET DATA
	client = Client('x','x')

	# valid intervals - 1m, 3m, 5m, 15m, 30m, 1h, 2h, 4h, 6h, 8h, 12h, 1d, 3d, 1w, 1M
	# get timestamp of earliest date data is available
	timestamp = client._get_earliest_valid_timestamp('BTCUSDT', '1d')

	# request historical candle (or klines) data
	bars = client.get_historical_klines('BTCUSDT', '1d', timestamp, limit=1000)

	# delete unwanted data - just keep date, open, high, low, close
	for line in bars:
	    del line[5:]

	# option 4 - create a Pandas DataFrame and export to CSV
	BTC_data = pd.DataFrame(bars, columns=['Date', 'BTC_24hOpenUSD', 'BTC_24hHighUSD', 'BTC_24hLowUSD', 'BTC_ClosingPriceUSD'])
	#-----------------------------------------------------------------------------------------------------


	#TRANSFORMACIÓN DATOS
	#Eliminamos las columnas que no utilizaremos en el estudio
	del BTC_data["BTC_24hHighUSD"]
	del BTC_data["BTC_24hLowUSD"]

	#Nos aseguramos que los datos deseados son numéricos
	BTC_data['BTC_ClosingPriceUSD'] = pd.to_numeric(BTC_data['BTC_ClosingPriceUSD'])
	BTC_data['BTC_24hOpenUSD'] = pd.to_numeric(BTC_data['BTC_24hOpenUSD'])

	#Nos aseguramos que el campo date es un datetime
	# BTC_data['Date'] = pd.to_datetime(BTC_data['Date'], format='%Y-%m-%d')
	BTC_data['Date'] = pd.to_datetime(BTC_data['Date'], unit='ms')

	#Nos aseguramos de los DataFrames están ordenados por fecha
	BTC_data = BTC_data.sort_values(by='Date', ascending=True)

	#Comprobación de que exite un único registro por fecha
	#Si hay algún registro sin fecha, lo crea e introduce los valores del día anterior, desde binance no he detectado ninguno
	num_rows = BTC_data.shape[0]-1
	i_i = BTC_data.index.min() #índice inicial
	d1 = BTC_data.loc[i_i, 'Date']
	d = 0 #contador de días
	i = i_i #contador de índices incluyendo nuevos registros
	n = i_i #contador de índices con solo los registros de las fechas encontradas

	while i <=num_rows:
	    comp_date = d1 + timedelta(days=d) #Fecha que tendremos que comprobar que existe
	    #Si existe la fecha
	    if BTC_data[BTC_data['Date']==comp_date]['Date'].size>0:
	        n+=1
	    #Si no existe la fecha
	    else:       
	        #Hallamos los datos del día anterior
	        closing_value = BTC_data.loc[n, 'BTC_ClosingPriceUSD']
	        open_value = BTC_data.loc[n, 'BTC_24hOpenUSD']
	        #Creamos el nuevo registro y le introducimos los datos del día anterior
	        # print("No existía registro de " + comp_date)
	        BTC_data.loc[num_rows+i]=[comp_date, closing_value, open_value]
	        BTC_data = BTC_data.sort_values(by='Date', ascending=True)
	        num_rows = BTC_data.shape[0]-1
	    i+=1
	    d+=1

	#Se reemplaza los valores por los BTC_ClosingPriceUSD del día anterior para asegurarnos que son correctos (descubrí que no lo eran, posiblemente por los llamados gaps en el precio de Bitcoin)
	#Despreciaremos el primer registro ya que no podremos encontrar el BTC_ClosingPriceUSD del día anterior
	BTC_data.loc[2:,"BTC_24hOpenUSD"] = BTC_data["BTC_ClosingPriceUSD"].shift(1)

	#Hallamos las fechas máximas y mínimas
	first_date = BTC_data['Date'].min()
	last_date = BTC_data['Date'].max()
	#-----------------------------------------------------------------------------------------------------


	#CALCULAR ALARMAS
	#Mejores medias móviles obtenidas en https://colab.research.google.com/drive/1z4aJ592OgnPNAWRgr2BHt8SrU7xVkfT8
	MA_a_max = 9
	MA_b_max = 23

	#Creamos las mejores medias móviles de para los precios de apertura y cierre
	BTC_data["BTC_Closing_BMA_a"] = BTC_data["BTC_ClosingPriceUSD"].rolling(window=MA_a_max,min_periods=0).mean()
	BTC_data["BTC_Closing_BMA_b"] = BTC_data["BTC_ClosingPriceUSD"].rolling(window=MA_b_max,min_periods=0).mean()
	BTC_data["BTC_24Open_BMA_a"] = BTC_data["BTC_24hOpenUSD"].rolling(window=MA_a_max,min_periods=0).mean()
	BTC_data["BTC_24Open_BMA_b"] = BTC_data["BTC_24hOpenUSD"].rolling(window=MA_b_max,min_periods=0).mean()

	#Creamos las señales de alarma de compra y venta
	#Si en un día la media móvil de 9 estaba por debajo de la de 23 y termina el día por encima, señal de compra = 1
	BTC_data["BTC_buy_alarm"] = [1 if Open_BMA_b > Open_BMA_a and Closing_BMA_a > Closing_BMA_b else 0 for (Open_BMA_a,Open_BMA_b,Closing_BMA_a,Closing_BMA_b) in zip(BTC_data['BTC_24Open_BMA_a'],BTC_data['BTC_24Open_BMA_b'],BTC_data['BTC_Closing_BMA_a'],BTC_data['BTC_Closing_BMA_b'])] 

	#Si en un día la media móvil de 9 estaba por arriba de la de 23 y termina el día por debajo, señal de venta = 1
	BTC_data["BTC_sell_alarm"] = [1 if Open_BMA_b < Open_BMA_a and Closing_BMA_a < Closing_BMA_b else 0 for (Open_BMA_a,Open_BMA_b,Closing_BMA_a,Closing_BMA_b) in zip(BTC_data['BTC_24Open_BMA_a'],BTC_data['BTC_24Open_BMA_b'],BTC_data['BTC_Closing_BMA_a'],BTC_data['BTC_Closing_BMA_b'])]  

	buy = BTC_data.loc[BTC_data.index[-2], "BTC_buy_alarm"]
	sell = BTC_data.loc[BTC_data.index[-2], "BTC_sell_alarm"]
	comparation_MA = BTC_data.loc[BTC_data.index[-2], "BTC_Closing_BMA_a"] - BTC_data.loc[BTC_data.index[-2], "BTC_Closing_BMA_b"]
	#-----------------------------------------------------------------------------------------------------


	#CREACIÓN MENSAJE
	last_close_price = BTC_data.loc[BTC_data.index[-2], "BTC_ClosingPriceUSD"]
	last_open_price = BTC_data.loc[BTC_data.index[-2], "BTC_24hOpenUSD"]
	last_variation = (last_close_price-last_open_price)/last_open_price*100
	last_close_price_format = "{:,}".format(last_close_price).replace(',','~').replace('.',',').replace('~','.')
	last_open_price_format = "{:,}".format(last_open_price).replace(',','~').replace('.',',').replace('~','.')

	if last_variation>=0:
	    last_variation_format = '<span style="color: MediumSeaGreen">+'+ str(round(last_variation,2)) + '%</span>'
	else:
	    last_variation_format = '<span style="color: red">' + str(round(last_variation,2)) + '%</span>'

	extra_text = "<ul>  <li>Open: " + str(last_open_price_format) + "$</li>  <li>Close: " + str(last_close_price_format) + "$</li>  <li>Change: " + last_variation_format +"</li></ul>"
	
	subject_buy = "BTC BUY ALARM"
	message_buy = '<h1 style="color:MediumSeaGreen;">HA SALTADO LA SEÑAL DE COMPRA DE BTC</h1>'
	subject_sell = "BTC SELL ALARM"
	message_sell = '<h1 style="color:red;">HA SALTADO LA SEÑAL DE VENTA DE BTC</h1>'
	subject_stay_in = "BTC alarm updated"
	message_stay_in = "<h2>PERMANECE DENTRO DEL MERCADO</h2>"
	subject_stay_out = "BTC alarm updated"
	message_stay_out = "<h2>PERMANECE FUERA DEL MERCADO</h2>"

	if buy == 1:
	  subject = subject_buy
	  text = message_buy
	elif sell == 1:
	  subject = subject_sell
	  text = message_sell
	elif comparation_MA > 0:
	  subject = subject_stay_in
	  text = message_stay_in
	else:
	  subject = subject_stay_out
	  text = message_stay_out

	text_inicial = "<p>Se han actualizado las señales de compra/venta.</p>"
	text = text_inicial + text + extra_text
	
	sender_mail = '*****@*****.**'
	receiver_mail = '*****@*****.**'

	message = Mail(
	    from_email=sender_mail,
	    to_emails=receiver_mail,
	    subject=subject,
	    html_content=text)
	try:
	    sg = SendGridAPIClient(os.environ.get('SENDGRID_API_KEY'))
	    response = sg.send(message)
	    print(response.status_code)
	    print(response.body)
	    print(response.headers)
	    return "Ha funcionado"
	except Exception as e:
	    print(e.message)
	    return "No ha funcionado"
Esempio n. 6
0
from binance.client import Client
import os
import configparser
import pandas as pd
api_key = os.environ.get('binance_api')
api_secret = os.environ.get('binance_secret')

client = Client(api_key, api_secret)
client.API_URL = 'https://api.binance.us/api'

tickers = client.get_account()
balance = client.get_asset_balance(asset='ETH')

# Getting earliest timestamp availble (on Binance)
earliest_timestamp = client._get_earliest_valid_timestamp(
    'ETHUSDT',
    '1d')  # Here "ETHUSDT" is a trading pair and "1d" is time interval
print(earliest_timestamp)

# Getting historical data (candle data or kline)
candle = client.get_historical_klines("ETHUSDT",
                                      "1d",
                                      earliest_timestamp,
                                      limit=1000)

print(candle[1])
#print(balance)

eth_df = pd.DataFrame(candle,
                      columns=[
                          'dateTime', 'open', 'high', 'low', 'close', 'volume',