Python bcolz.carray方法代码示例

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def perform_val(embedding_size, batch_size, model,
                carray, issame, nrof_folds=10, is_ccrop=False, is_flip=True):
    """perform val"""
    embeddings = np.zeros([len(carray), embedding_size])

    for idx in tqdm.tqdm(range(0, len(carray), batch_size)):
        batch = carray[idx:idx + batch_size]
        batch = np.transpose(batch, [0, 2, 3, 1]) * 0.5 + 0.5
        if is_ccrop:
            batch = ccrop_batch(batch)
        if is_flip:
            fliped = hflip_batch(batch)
            emb_batch = model(batch) + model(fliped)
            embeddings[idx:idx + batch_size] = l2_norm(emb_batch)
        else:
            batch = ccrop_batch(batch)
            emb_batch = model(batch)
            embeddings[idx:idx + batch_size] = l2_norm(emb_batch)

    tpr, fpr, accuracy, best_thresholds = evaluate(
        embeddings, issame, nrof_folds)

    return accuracy.mean(), best_thresholds.mean() 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def get_val_pair(path, name):
    carray = bcolz.carray(rootdir=os.path.join(path, name), mode='r')
    issame = np.load('{}/{}_list.npy'.format(path, name))

    return carray, issame 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def save_array(fname, arr):
    c=bcolz.carray(arr, rootdir=fname, mode='w')
    c.flush() 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def bcolz_save(path, np_array):
    c = bcolz.carray(np_array, rootdir=path, mode='w')
    c.flush()
    print("Saved to " + path) 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def write(self, rootdir):
        """
        Write the metadata to a JSON file in the rootdir.

        Values contained in the metadata are:
        first_trading_day : string
            'YYYY-MM-DD' formatted representation of the first trading day
             available in the dataset.
        minute_index : list of integers
             nanosecond integer representation of the minutes, the enumeration
             of which corresponds to the values in each bcolz carray.
        ohlc_ratio : int
             The factor by which the pricing data is multiplied so that the
             float data can be stored as an integer.
        """
        metadata = {
            'first_trading_day': str(self.first_trading_day.date()),
            'minute_index': self.minute_index.asi8.tolist(),
            'market_opens': self.market_opens.values.
            astype('datetime64[m]').
            astype(int).tolist(),
            'market_closes': self.market_closes.values.
            astype('datetime64[m]').
            astype(int).tolist(),
            'ohlc_ratio': self.ohlc_ratio,
        }
        with open(self.metadata_path(rootdir), 'w+') as fp:
            json.dump(metadata, fp) 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def _open_minute_file(self, field, sid):
        sid = int(sid)

        try:
            carray = self._carrays[field][sid]
        except KeyError:
            carray = self._carrays[field][sid] = \
                bcolz.carray(rootdir=self._get_carray_path(sid, field),
                             mode='r')

        return carray 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def _open_minute_file(self, field, sid):
        try:
            carray = self._carrays[field][sid]
        except KeyError:
            carray = self._carrays[field][sid] = \
                bcolz.carray(rootdir=self._get_carray_path(sid, field),
                             mode='r')

        return carray 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def _filtered_index(self, instrument):
        # TODO 确认是否跳过日内涨跌停
        if instrument not in self._index_skip_suspending:
            carray = self._open_minute_file("close", instrument)
            sub_index = bcolz.eval("carray != 0", vm="numexpr")
            index = self._minute_index[:len(sub_index)][sub_index[:]]
            self._index_skip_suspending[instrument] = index
        return self._index_skip_suspending[instrument] 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def unique(self, col_or_col_list):
        """
        Return a list of unique values of a column or a list of lists of column list

        :param col_or_col_list: a column or a list of columns
        :return:
        """

        if isinstance(col_or_col_list, list):
            col_is_list = True
            col_list = col_or_col_list
        else:
            col_is_list = False
            col_list = [col_or_col_list]

        output = []

        for col in col_list:

            if self.auto_cache or self.cache_valid(col):
                # create factorization cache
                if not self.cache_valid(col):
                    self.cache_factor([col])

                # retrieve values from existing disk-based factorization
                col_values_rootdir = self[col].rootdir + '.values'
                carray_values = bcolz.carray(rootdir=col_values_rootdir, mode='r')
                values = list(carray_values)
            else:
                # factorize on-the-fly
                _, values = ctable_ext.factorize(self[col])
                values = values.values()

            output.append(values)

        if not col_is_list:
            output = output[0]

        return output 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def factorize_groupby_cols(self, groupby_cols):
        """
        factorizes all columns that are used in the groupby
        it will use cache carrays if available
        if not yet auto_cache is valid, it will create cache carrays

        """
        # first check if the factorized arrays already exist
        # unless we need to refresh the cache
        factor_list = []
        values_list = []

        # factorize the groupby columns
        for col in groupby_cols:

            if self.auto_cache or self.cache_valid(col):
                # create factorization cache if needed
                if not self.cache_valid(col):
                    self.cache_factor([col])

                col_rootdir = self[col].rootdir
                col_factor_rootdir = col_rootdir + '.factor'
                col_values_rootdir = col_rootdir + '.values'
                col_carray_factor = \
                    bcolz.carray(rootdir=col_factor_rootdir, mode='r')
                col_carray_values = \
                    bcolz.carray(rootdir=col_values_rootdir, mode='r')
            else:
                col_carray_factor, values = ctable_ext.factorize(self[col])
                col_carray_values = \
                    bcolz.carray(np.fromiter(values.values(), dtype=self[col].dtype))

            factor_list.append(col_carray_factor)
            values_list.append(col_carray_values)

        return factor_list, values_list 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def _int_array_hash(input_list):
        """
        A function to calculate a hash value of multiple integer values, not used at the moment

        Parameters
        ----------
        input_list

        Returns
        -------

        """

        list_len = len(input_list)
        arr_len = len(input_list[0])
        mult_arr = np.full(arr_len, 1000003, dtype=np.long)
        value_arr = np.full(arr_len, 0x345678, dtype=np.long)

        for i, current_arr in enumerate(input_list):
            index = list_len - i - 1
            value_arr ^= current_arr
            value_arr *= mult_arr
            mult_arr += (82520 + index + index)

        value_arr += 97531
        result_carray = bcolz.carray(value_arr)
        del value_arr
        return result_carray 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def unadjusted_window(self, fields, start_dt, end_dt, sids):
        """
        Parameters
        ----------
        fields : list of str
           'open', 'high', 'low', 'close', or 'volume'
        start_dt: Timestamp
           Beginning of the window range.
        end_dt: Timestamp
           End of the window range.
        sids : list of int
           The asset identifiers in the window.

        Returns
        -------
        list of np.ndarray
            A list with an entry per field of ndarrays with shape
            (sids, minutes in range) with a dtype of float64, containing the
            values for the respective field over start and end dt range.
        """
        # TODO: Handle early closes.
        start_idx = self._find_position_of_minute(start_dt)
        end_idx = self._find_position_of_minute(end_dt)

        results = []

        shape = (len(sids), (end_idx - start_idx + 1))

        for field in fields:
            if field != 'volume':
                out = np.full(shape, np.nan)
            else:
                out = np.zeros(shape, dtype=np.uint32)

            for i, sid in enumerate(sids):
                carray = self._open_minute_file(field, sid)
                values = carray[start_idx:end_idx + 1]
                where = values != 0
                out[i, where] = values[where]
            if field != 'volume':
                out *= self._ohlc_inverse
            results.append(out)
        return results 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def raw_history_bars(self, instrument, start_dt=None, end_dt=None, length=None, fields=None, skip_suspended=True):
        """

        Parameters
        ----------
        instrument
        start_dt
        end_dt
        length
        fields
        skip_suspended

        Returns
        -------

        """
        start_idx, end_idx = self.get_dt_slice(instrument, start_dt, end_dt, length, skip_suspended)
        if fields is None:
            fields_ = self.FIELDS
        elif isinstance(fields, six.string_types):
            fields_ = [fields]
        else:
            fields_ = [field for field in fields if field != "datetime"]
        num_minutes = end_idx - start_idx
        types = {f: self._converter.field_type(f, np.float64) for f in fields_}
        dtype = np.dtype([("datetime", np.uint64)] +
                         [(f, self._converter.field_type(f, np.float64)) for f in fields_])
        shape = (num_minutes,)
        result = np.empty((num_minutes,), dtype=dtype)
        for field in fields_:
            if field != 'volume':
                out = np.full(shape, np.nan, dtype=types[field])
            else:
                out = np.zeros(shape, dtype=types[field])
            carray = self._open_minute_file(field, instrument)
            values = carray[start_idx: end_idx]
            where = values != 0
            if field != 'volume':
                out[:len(where)][where] = values[where] * self._ohlc_ratio_inverse_for_sid(instrument)
            else:
                out[:len(where)][where] = values[where]
            result[field] = out
        result["datetime"] = list(map(convert_dt_to_int, self._minute_index[start_idx: end_idx].to_pydatetime()))
        result = result if fields is None else result[fields]
        return self.numba_loops_dropna(result) if skip_suspended else self.numba_loops_ffill(result) 

# 需要导入模块: import bcolz [as 别名]
# 或者: from bcolz import carray [as 别名]
def cache_factor(self, col_list, refresh=False):
        """
        Existing todos here are: these should be hidden helper carrays
        As in: not normal columns that you would normally see as a user

        The factor (label index) carray is as long as the original carray
        (and the rest of the table therefore)
        But the (unique) values carray is not as long (as long as the number
        of unique values)

        :param col_list:
        :param refresh:
        :return:
        """

        if not self.rootdir:
            raise TypeError('Only out-of-core ctables can have '
                            'factorization caching at the moment')

        if not isinstance(col_list, list):
            col_list = [col_list]

        if refresh:
            kill_list = [x for x in os.listdir(self.rootdir) if '.factor' in x or '.values' in x]
            for kill_dir in kill_list:
                rm_file_or_dir(os.path.join(self.rootdir, kill_dir))

        for col in col_list:

            # create cache if needed
            if refresh or not self.cache_valid(col):
                # todo: also add locking mechanism here

                # create directories
                col_rootdir = self[col].rootdir
                col_factor_rootdir = col_rootdir + '.factor'
                col_factor_rootdir_tmp = tempfile.mkdtemp(prefix='bcolz-')
                col_values_rootdir = col_rootdir + '.values'
                col_values_rootdir_tmp = tempfile.mkdtemp(prefix='bcolz-')

                try:
                    # create factor
                    carray_factor = \
                        bcolz.carray([], dtype='int64', expectedlen=self.size,
                                     rootdir=col_factor_rootdir_tmp, mode='w')
                    _, values = \
                        ctable_ext.factorize(self[col], labels=carray_factor)
                    carray_factor.flush()
                finally:
                    rm_file_or_dir(col_factor_rootdir, ignore_errors=True)
                    shutil.move(col_factor_rootdir_tmp, col_factor_rootdir)

                try:
                    # create values
                    carray_values = \
                        bcolz.carray(np.fromiter(values.values(), dtype=self[col].dtype),
                                     rootdir=col_values_rootdir_tmp, mode='w')
                    carray_values.flush()
                finally:
                    rm_file_or_dir(col_values_rootdir, ignore_errors=True)
                    shutil.move(col_values_rootdir_tmp, col_values_rootdir)