diff options
| -rw-r--r-- | python/load_globeop_report.py | 30 |
1 files changed, 12 insertions, 18 deletions
diff --git a/python/load_globeop_report.py b/python/load_globeop_report.py index 6fd998b9..03297caa 100644 --- a/python/load_globeop_report.py +++ b/python/load_globeop_report.py @@ -3,18 +3,17 @@ import os import pandas as pd from itertools import chain from dates import bus_day -import pdb -def get_globs(fname, years=['2013', '2014', '2015', '2016']): - basedir = '/home/share/Daily' +def get_globs(fname, years=['2013', '2014', '2015', '2016', '2017']): + basedir = '/home/serenitas/Daily' globs = [iglob(os.path.join(basedir, year, "{0}_*/{0}*/Reports/{1}.csv".format(year, fname))) for year in years] for year in years[-2:]: globs.append(iglob(os.path.join(basedir, - '{0}-*/Reports/{1}.csv'.format(year, - fname)))) + '{0}-*/Reports/{1}.csv'.format(year, + fname)))) return globs def valuation_reports(): @@ -25,18 +24,18 @@ def valuation_reports(): except ValueError: date = pd.Timestamp(f.split('/')[4]) - if date>=pd.Timestamp('2013-02-06'): + if date >= pd.Timestamp('2013-02-06'): newdf = pd.read_csv(f, parse_dates=['KnowledgeDate','PeriodEndDate']) else: newdf = pd.read_csv(f) newdf['KnowledgeDate'] = date newdf['PeriodEndDate'] = date - bus_day if newdf.empty or ('PeriodEndDate' in df and \ - not df[df.PeriodEndDate == newdf.PeriodEndDate.iat[0]].empty): + not df[df.PeriodEndDate == newdf.PeriodEndDate.iat[0]].empty): continue df = df.append(newdf) del df['AccountingPeriod'] - for col in ['Strat','InvCcy','Fund','Port']: + for col in ['Strat', 'InvCcy', 'Fund', 'Port']: df[col] = df[col].astype('category') df.to_hdf('globeop.hdf', 'valuation_report', format='table', complib='blosc') @@ -48,8 +47,6 @@ def pnl_reports(): except ValueError: date = pd.Timestamp(f.split('/')[4]) date = date - bus_day - if date in df: - print(date) df[date] = pd.read_csv(f) df = pd.concat(df, names=['date', 'to_drop']) df.reset_index(level='to_drop', drop=True, inplace=True) @@ -58,21 +55,18 @@ def pnl_reports(): df[col] = df[col].astype('category') df.to_hdf('globeop.hdf', 'pnl', format='table', complib='blosc') -def ts(s): - return pd.Timestamp(s) - def monthly_pnl_bycusip(df, strats): df = df[(df.Strat.isin(strats)) & (df.CustAcctName=='V0NSCLMAMB')] pnl_cols = ['BookUnrealMTM', 'BookRealMTM', 'BookRealIncome', 'BookUnrealIncome', 'TotalBookPL'] - return df.groupby('InvId').resample('M', 'last')[['MTD '+col for col in pnl_cols]] + return df.groupby('InvId').resample('M').last()[['MTD '+col for col in pnl_cols]] if __name__=='__main__': - #valuation_reports() - #pnl_reports() + valuation_reports() + pnl_reports() df_val = pd.read_hdf('globeop.hdf', 'valuation_report') df_pnl = pd.read_hdf('globeop.hdf', 'pnl') - nav = df_val[df_val.Fund=='SERCGMAST'].groupby('PeriodEndDate')['EndBookNAV'].sum() + nav = df_val[df_val.Fund == 'SERCGMAST'].groupby('PeriodEndDate')['EndBookNAV'].sum() subprime_strats = ['MTG_GOOD', 'MTG_RW', 'MTG_IO','MTG_THRU', 'MTG_B4PR'] clo_strats = ['CLO_BBB', 'CLO_AAA', 'CLO_BB20'] @@ -83,6 +77,6 @@ if __name__=='__main__': #df_monthly.loc[idx[ts('2015-01-01'):ts('2015-01-31'),:],:] # clo = df_pnl[df_pnl.Strat.isin(clo_strats)] - # clo_monthly_pnl = clo.groupby(level=0).sum()['MTD TotalBookPL'].resample('M', how='last') + # clo_monthly_pnl = clo.groupby(level=0).sum()['MTD TotalBookPL'].resample('M').last() # clo.groupby(level=0).sum()['2015-12-01':'2015-12-31'] |