forked from pandas-dev/pandas
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathtimeseries.py
129 lines (95 loc) · 4.06 KB
/
timeseries.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
from vbench.api import Benchmark
from datetime import datetime
common_setup = """from pandas_vb_common import *
from datetime import timedelta
N = 100000
try:
rng = date_range('1/1/2000', periods=N, freq='min')
except NameError:
rng = DateRange('1/1/2000', periods=N, offset=datetools.Minute())
def date_range(start=None, end=None, periods=None, freq=None):
return DateRange(start, end, periods=periods, offset=freq)
ts = Series(np.random.randn(N), index=rng)
"""
#----------------------------------------------------------------------
# Lookup value in large time series, hash map population
setup = common_setup + """
rng = date_range('1/1/2000', periods=1500000, freq='s')
ts = Series(1, index=rng)
"""
stmt = "ts[ts.index[len(ts) // 2]]; ts.index._cleanup()"
timeseries_large_lookup_value = Benchmark(stmt, setup,
start_date=datetime(2012, 1, 1))
#----------------------------------------------------------------------
# Test slice minutely series
timeseries_slice_minutely = Benchmark('ts[:10000]', common_setup)
#----------------------------------------------------------------------
# Test conversion
setup = common_setup + """
"""
timeseries_1min_5min_ohlc = Benchmark("ts[:10000].convert('5min', how='ohlc')",
common_setup)
timeseries_1min_5min_mean = Benchmark("ts[:10000].convert('5min', how='mean')",
common_setup)
#----------------------------------------------------------------------
# Irregular alignment
setup = common_setup + """
lindex = np.random.permutation(N)[:N // 2]
rindex = np.random.permutation(N)[:N // 2]
left = Series(ts.values.take(lindex), index=ts.index.take(lindex))
right = Series(ts.values.take(rindex), index=ts.index.take(rindex))
"""
timeseries_add_irregular = Benchmark('left + right', setup)
#----------------------------------------------------------------------
# Sort large irregular time series
setup = common_setup + """
N = 100000
rng = date_range('1/1/2000', periods=N, freq='s')
rng = rng.take(np.random.permutation(N))
ts = Series(np.random.randn(N), index=rng)
"""
timeseries_sort_index = Benchmark('ts.sort_index()', setup,
start_date=datetime(2012, 4, 1))
#----------------------------------------------------------------------
# Shifting, add offset
setup = common_setup + """
rng = date_range('1/1/2000', periods=10000, freq='T')
"""
datetimeindex_add_offset = Benchmark('rng + timedelta(minutes=2)', setup,
start_date=datetime(2012, 4, 1))
setup = common_setup + """
N = 10000
rng = date_range('1/1/1990', periods=N, freq='53s')
ts = Series(np.random.randn(N), index=rng)
dates = date_range('1/1/1990', periods=N * 10, freq='5s')
"""
timeseries_asof_single = Benchmark('ts.asof(dates[0])', setup,
start_date=datetime(2012, 4, 27))
timeseries_asof = Benchmark('ts.asof(dates)', setup,
start_date=datetime(2012, 4, 27))
setup = setup + 'ts[250:5000] = np.nan'
timeseries_asof_nan = Benchmark('ts.asof(dates)', setup,
start_date=datetime(2012, 4, 27))
#----------------------------------------------------------------------
# Time zone stuff
setup = common_setup + """
rng = date_range('1/1/2000', '3/1/2000', tz='US/Eastern')
"""
timeseries_timestamp_tzinfo_cons = \
Benchmark('rng[0]', setup, start_date=datetime(2012, 5, 5))
#----------------------------------------------------------------------
# Resampling period
setup = common_setup + """
rng = period_range('1/1/2000', '1/1/2001', freq='T')
ts = Series(np.random.randn(len(rng)), index=rng)
"""
timeseries_period_downsample_mean = \
Benchmark("ts.resample('D', how='mean')", setup,
start_date=datetime(2012, 4, 25))
setup = common_setup + """
rng = date_range('1/1/2000', '1/1/2001', freq='T')
ts = Series(np.random.randn(len(rng)), index=rng)
"""
timeseries_timestamp_downsample_mean = \
Benchmark("ts.resample('D', how='mean')", setup,
start_date=datetime(2012, 4, 25))