def bdh(tickers,
flds=None,
start_date=None,
end_date='today',
adjust=None,
**kwargs) -> pd.DataFrame:
"""
Bloomberg historical data
Args:
tickers: ticker(s)
flds: field(s)
start_date: start date
end_date: end date - default today
adjust: `all`, `dvd`, `normal`, `abn` (=abnormal), `split`, `-` or None
exact match of above words will adjust for corresponding events
Case 0: `-` no adjustment for dividend or split
Case 1: `dvd` or `normal|abn` will adjust for all dividends except splits
Case 2: `adjust` will adjust for splits and ignore all dividends
Case 3: `all` == `dvd|split` == adjust for all
Case 4: None == Bloomberg default OR use kwargs
**kwargs: overrides
Returns:
pd.DataFrame
"""
logger = logs.get_logger(bdh, **kwargs)
if flds is None: flds = ['Last_Price']
e_dt = utils.fmt_dt(end_date, fmt='%Y%m%d')
if start_date is None:
start_date = pd.Timestamp(e_dt) - pd.Timedelta(weeks=8)
s_dt = utils.fmt_dt(start_date, fmt='%Y%m%d')
request = process.create_request(
service='//blp/refdata',
request='HistoricalDataRequest',
**kwargs,
)
process.init_request(request=request,
tickers=tickers,
flds=flds,
start_date=s_dt,
end_date=e_dt,
adjust=adjust,
**kwargs)
logger.debug(f'Sending request to Bloomberg ...\n{request}')
conn.send_request(request=request, **kwargs)
res = pd.DataFrame(process.rec_events(process.process_hist, **kwargs))
if kwargs.get('raw', False): return res
if res.empty or any(fld not in res for fld in ['ticker', 'date']):
return pd.DataFrame()
return (res.set_index(['ticker', 'date']).unstack(level=0).rename_axis(
index=None, columns=[None, None]).swaplevel(
0, 1, axis=1).reindex(columns=utils.flatten(tickers),
level=0).reindex(columns=utils.flatten(flds),
level=1))
def bdp_bds_cache(func, tickers, flds, **kwargs) -> ToQuery:
"""
Find cached `BDP` / `BDS` queries
Args:
func: function name - bdp or bds
tickers: tickers
flds: fields
**kwargs: other kwargs
Returns:
ToQuery(ticker, flds, kwargs)
"""
cache_data = []
log_level = kwargs.get('log', logs.LOG_LEVEL)
logger = logs.get_logger(bdp_bds_cache, level=log_level)
has_date = kwargs.pop('has_date', func == 'bds')
cache = kwargs.get('cache', True)
tickers = utils.flatten(tickers)
flds = utils.flatten(flds)
loaded = pd.DataFrame(data=0, index=tickers, columns=flds)
for ticker, fld in product(tickers, flds):
data_file = storage.ref_file(
ticker=ticker,
fld=fld,
has_date=has_date,
cache=cache,
ext='pkl',
**{k: v
for k, v in kwargs.items() if k not in EXC_COLS})
if not files.exists(data_file): continue
logger.debug(f'reading from {data_file} ...')
cache_data.append(pd.read_pickle(data_file))
loaded.loc[ticker, fld] = 1
to_qry = loaded.where(loaded == 0)\
.dropna(how='all', axis=1).dropna(how='all', axis=0)
return ToQuery(tickers=to_qry.index.tolist(),
flds=to_qry.columns.tolist(),
cached_data=cache_data)