DSC 201: Data Analysis & Visualization
[Pages:13]DSC 201: Data Analysis & Visualization
Time Series
Dr. David Koop
D. Koop, DSC 201, Fall 2016
Dates and Times
? What is time to a computer? - Can be stored as seconds since Unix Epoch (January 1st, 1970)
? Often useful to break down into minutes, hours, days, months, years...
? Lots of different ways to write time: - How could you write "November 29, 2016"? - European vs. American ordering...
? What about time zones?
D. Koop, DSC 201, Fall 2016
2
Python Support for Time
? The datetime package - Has date, time, and datetime classes
- .now() method: the current datetime - Can access properties of the time (year, month, seconds, etc.)
? Converting from strings to datetimes:
- datetime.strptime: good for known formats - dateutil.parser.parse: good for unknown formats ? Converting to strings
- str(dt) or dt.strftime() ? Differences between times
- datetime.timedelta: can get number of days/hours/etc. - deal with issues with different length months, etc.
D. Koop, DSC 201, Fall 2016
3
Pandas Support for Datetime
? pd.to_datetime: - convenience method - can convert an entire column to datetime
? Has a NaT to indicate a missing time value (like NaN but for timestamps)
? Stores in a numpy.datetime64 format ? pd.Timestamp: a wrapper for the datetime64 objects ? Can use time as an index ? Accessing a particular time or checking equivalence allows any
string that can be interpreted as a date: - ts['1/10/2011'] or ts['20110110']
D. Koop, DSC 201, Fall 2016
4
In [87]: hour Out[87]:
GYoeuncaenrdaetfiinnega mDualtitpele Rof aannogffesest by passing an integer:
In [88]: four_hours = Hour(4)
? index = pd.date_range('4/1/2012', '6/1/2012')
? CaInng[e89n]e:rafoteur_bhaousresd on a number of periods as well
Out[89]:
- index = pd.date_range('4/1/2012', periods=20)
?IinnFsrmteeqaodsuteuanspicnpygli(acfasrttieroiqnn)gs,cayolionaustrlowiklsoeuh'lHdo'wnoertvh'e4erHrn'a.enPedguetttoiinsegdxapinvliidicniettldeygecrrebaetfeoroentehoe fbtahseesfereoqb cr-eaCtoesdaesmfuolrtipsplee: cifying this (e.g. 4h, D, M)
- In [90]: pd.date_range('1/1/2000', '1/3/2000 23:59', freq='4h')
Out[90]: [2000-01-01 00:00:00, ..., 2000-01-03 20:00:00] Length: 18, Freq: 4H, Timezone: None
M-aCnyanofafslestos mcainx btheecmom: b'2inhe3d0tmog'ether by addition:
In [91]: Hour(2) + Minute(30) Out[91]:
Similarly, you can pass frequency strings like '2h30min' which will effectively be p to the same expression:
D. Koop,IDnS[C922]0:1, pFda.lld2a0t1e6_range('1/1/2000', periods=10, freq='1h30min')
5
Assignment 5
? ? Aggregation, resampling, and visualization of time series data ? Last assignment, but likely challenging...start now!
D. Koop, DSC 201, Fall 2016
6
Shifting dates with offsets
[102]: ts.shift(1, freq='90T')
[10T2T]h:iempeadnedaltsadate offsets can also be used with datetime or Ti
00--0012?--C3219oImn0011p[::u133te0003::d]00i:f00fefrerno-cm00e..sp08ab63ne68dt76aw43s89e.etnsedraiteess.offsets import Day, MonthEnd
00--0034?--L33i10vIen00s11i[::n133d000a4::]t00:e00tniomw--e00=m..15od11da77ut37lee89t85ime(2011,
pe: ?fddlioiIfafntff6.[4=d1a0py5as]r:sen_odwat+e(3"1*
Jan 2017")
Day()
11, 17)
- datetime.now().date()
? AlsOouta[p1d0.5]T:imTeidmeelsttaamopb(je'c2t01th1a-t1t1a-k2e0s0tr0in:g0s0::00')
datIefsy-woduiatthaedotfdifmsaeent.snaonwc(h).odraetde(o)ffs+etpldi.kTeiMmeodnetlhtEan(d",4thdeayfisr"s)t incremen
ndtao?sAtdhlsaeot,enReooxfllftdsdaettaestsecuasnicncgaoalrsndocihnbogeretduosoetffhdseewtsfriethqudeantceytirmueleo:r Timestamp
[103]: Ifnro[m10p6a]n:danso.wts+erMioenst.hoEfnfds(e)ts import Day, MonthEnd
Out[106]: Timestamp('2011-11-30 00:00:00')
[104]: now = datetime(2011, 11, 17)
In [107]: now + MonthEnd(2)
[105]: Onuotw[1+073]:* TDiamye(s)tamp('2011-12-31 00:00:00')
[10A5]n:chTiomreesdtaomfpf(s'e2t0s1c1a-1n1-e2x0pl0i0c:it0l0y:0"0r'o)ll" dates forward or back
D. Koop, DSC 201, Fall 2016
7
Time Zones
? Why? ? Coordinated Universal Time (UTC) is the standard time (basically
equivalent to Greenwich Mean Time (GMT) ? Other time zones are UTC +/- a number in [1,12] ? Dartmouth is UTC-5 (aka US/Eastern)
D. Koop, DSC 201, Fall 2016
8
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related download
- pandas powerful python data analysis toolkit
- pyarrow documentation
- reading and writing data with pandas
- 10 minutes to pandas
- pandas validation documentation
- python tutorial for cse 446 university of washington
- assign column to dataframe pandas
- dsc 201 data analysis visualization
- 1 pandas 4 time series
- 12 pandas 4 time series brigham young university
Related searches
- data analysis questions examples
- data analysis research paper example
- data analysis method
- data analysis methods examples
- data analysis methods in research
- types of data analysis methods
- data analysis in research methodology
- data analysis in research pdf
- examples of data analysis paper
- data analysis techniques for research
- data analysis and interpretation pdf
- data analysis tools