ASA Connect

 View Only
Expand all | Collapse all

Ancient Datasets

  • 1.  Ancient Datasets

    Posted 05-15-2018 15:10
    I know the ancient Egyptians collected endless data on the flooding of the Nile. Is this data available online? Has it been used to teach time series methods?

    Glenn Cummings


  • 2.  RE: Ancient Datasets

    Posted 05-16-2018 16:09
    Hi Folks:  While it is possible that such data exists there will be some serious format issues and I do not read old Egyptian.  However the address I have included below may well have and interesting reply.  They are the reincarnation of the Ancient Library at Alexandria.  Good Luck.  Lawrence Lessner

     

    Maha Abbas

    Head of BAIFA Unit

    Bibliotheca Alexandrina Information for Africa (BAIFA)

    Networks Section

    Networks & Outreach department

    Bibliotheca Alexandrina

    Chatby 21526

    Alexandria, EGYPT

    Tel: +203 4839999  - Ext.: 2601

    Mobile: +2 0100 721 5316

    Email: maha.abbas@bibalex.org

    Website: www.bibalex.org







  • 3.  RE: Ancient Datasets

    Posted 05-16-2018 16:14
    Estimates of of the annual flow of the Nile at Aswan, 1871–1970 (Cobb, 1978, Table 1, p. 249), are available in the R package datasets (which is part of the base R distribution) as a data object called Nile. 

    This dataset has been used to demonstrate change points and the Hurst phenomenon (memory, aka serial correlation in a river). See the R package HKprocess for an autocorrelation example with the Nile data.


    Karen Ryberg





  • 4.  RE: Ancient Datasets

    Posted 05-16-2018 18:45
    Thanks, That is exactly what I was looking for, and easy to access since it is part of the R installation. I may still look at the full record (even with the gaps), but that really is too much information for an undergraduate statistics demonstration.

    ------------------------------
    Glenn Cummings
    ------------------------------



  • 5.  RE: Ancient Datasets

    Posted 05-18-2018 08:14
    An earlier (and I think more extensive) source for data on the Nile is

    @book{Popper:1951,
    ?? title={The Cairo Nilometer: Studies in Ibn Taghr{\^\i} Bird{\^\i}'s
    Chronicles of Egypt: I},
    ?? author={Popper, W.},
    ?? number={v. 12},
    ?? lccn={a51009495},
    ?? series={Publications in Semitic Philology},
    ?? url={https://books.google.ca/books?id=CrESAAAAIAAJ},
    ?? year={1951},
    ?? publisher={University of California Press}
    }

    Another source for historical datasets is my R package HistData,
    https://cran.r-project.org/web/packages/HistData/index.html

    -Michael

    --
    Michael Friendly Email: friendly AT yorku DOT ca
    Professor, Psychology Dept. & Chair, ASA Statistical Graphics Section
    York University Voice: 416 736-2100 x66249 Fax: 416 736-5814
    4700 Keele Street Web:http://www.datavis.ca
    Toronto, ONT M3J 1P3 CANADA




  • 6.  RE: Ancient Datasets

    Posted 05-18-2018 10:40
    A part of the Popper data set (annual minimum water levels, 622-1284) is available online at http://mldata.org/repository/data/viewslug/nile-water-level/

    Even older data, back to 3000 B.C., is described and graphed in Bell, B. (1970), The oldest records of the Nile floods, The Geographical Journal, 136(4), 569-573. doi:10.2307/179618.



    ------------------------------
    J. R. M. Hosking
    jrmhosking@gmail.com
    ------------------------------



  • 7.  RE: Ancient Datasets

    Posted 05-16-2018 16:28
    Glenn
    Very interesting.  THe egyptians were very good in recording everything. Also, there predictions were quite accurate

    You might want to go to the new library of alexandria

    Contact Ask a librarian.  If you cannot obtain a source, let me know and I will reach out to my friends there. It would be amazing to do time series analysis on data that are 3000 years old

    The data were certainly collected, but with the destruction of the library, virtual all of the scrolls were lost 

    ron

    "Health is a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity". (WHO)

    Ronald LaPorte,  Ph.D.
    Professor Emeritus, former WHO Collaborating Centre
    Department of Epidemiology
    Graduate School of Public Health
    130 Desoto Street
    University of Pittsburgh
    Pittsburgh, PA  15261
    USA

    Telephone home 724 934 9023 
    Mobile  724 759 3283

    Global Health, So Near, So Far

    Supercourse Project - www.pitt.edu/~super1 
    Science Supercourse - ssc.bibalex.org 
    Research Methods Library of Alexandria - http://ssc.bibalex.org/helpdesk/introduction.jsf
    Bibliotheca Alexandrina (BA) African Networks - http://afn.bibalex.org/GeneralPortal.aspx
    Central Asian Journal of Global Health -  cajgh.pitt.edu/





  • 8.  RE: Ancient Datasets

    Posted 05-17-2018 02:35
    You may want to ask at OpenData.StackExchange: Open Data Stack Exchange. There doesn't yet seem to be anything containing "Nile". Asking there would leave a searchable footprint for later generations.

    Open Data Stack Exchange remove preview
    Open Data Stack Exchange
    Q&A for developers and researchers interested in open data
    View this on Open Data Stack Exchange >


    ------------------------------
    Stephan Kolassa
    ------------------------------



  • 9.  RE: Ancient Datasets

    Posted 05-18-2018 15:52
    Hey,

    It looks like that data is available - it's used in the following publication:

    Garnett, R., Osborne, M., and Roberts, S. Sequential Bayesian prediction in the presence of changepoints. In Proceedings of the 26th International Conference on Machine Learning, pp. 345–352, Montreal, QC, June 2009.

    Link: http://www.robots.ox.ac.uk/~parg/pubs/changepoint.pdf


    They cite the following related publication:

    Whitcher, B., Byers, S., Guttorp, P., & Percival, D. (2002). Testing for homogeneity of variance in time series: Long memory, wavelets and the Nile River. Water Resources Research, 38, 10–1029.


    Regards,

    ------------------------------
    Glen Wright Colopy
    DPhil Student
    University of Oxford
    ------------------------------



  • 10.  RE: Ancient Datasets

    Posted 05-21-2018 15:19

    Jonathan –

    Your reference to the earlier paper by Bell (1970) is really useful, on an issue that is quite intriguing, compilation of time series which span several millennia.  

    Most of the research that has been done on the Nile is concerned with the predictability of the river flows. The Nile data set is interesting because it appears to exhibit long cycles, first identified in the seminal work of Hurst (1951).  So forecasting the Nile is a topic clearly worthy of investigation.

    Some years back, I was speculating about the possibility that the Nile cycles could have influenced the rise and fall of dynasties in Ancient Egypt. Egypt is famous for its bureaucratic, centralized monarchy. But this was true less than half the time. The three periods of political stability and centralization, the Old Kingdom, Middle Kingdom and New Kingdom, collectively span about 1200 years. The early dynastic period, prior to the Old Kingdom, the two intermediate periods, and the long period of decline beginning around 1100 BC, were characterized by civil wars and political disintegration.  

    I am wondering if the political transitions could have been influenced by long cycles in the Nile. The intuition of course is that Egypt's wealth depended heavily on the Nile, so that long periods of drought would have been more likely to undercut tax revenues, weakening the central monarchy.

    Has anyone ever looked at the earlier data, and is it sufficiently detailed to analyze this issue?

     

     

    Even older data, back to 3000 B.C., is described and graphed in Bell, B. (1970), The oldest records of the Nile floods, The Geographical Journal, 136(4), 569-573. doi:10.2307/179618.



    ------------------------------
    Gordon Reikard
    ------------------------------