Sinds 2002 werkzaam bij bibliotheken van de universiteit van Amsterdam, daarvoor in Ecuador bij proyecto Paramo.
Every month CAS offers us to download very usefull and
detailed Scifinder statistics. The IPDetail file is a
comma-separated values (CSV) file. It has 10 columns (March
2012) with use(r) information. Most useful data is the date,
time, minutes used (or busies) and login id (or one of the
other unique user identifiers like e-mail address etc.)
The file logs the start date and start time of the user and
the minutes of Scifinder use.
The next step in analyzing the log file is converting the user
data into a time series.Conversion of the non-equally spaced
values of the log into a list based on equally spaced
timevalues (minutes) and number of concurrent users (and
busies).
The program we use is available at the bottom and
iscalled:
"SF IPDetail converter"
Use the original IPDetail csv provided by cas for
download.
You will get:
Two files and a graph.
File A: detail concurrent user file
File B: converted ipdetail file
Graph : visual representation of file A. Flot jquery library
is used to draw the data.
The above graph is based on the converted data in the
IPDetail file. The graph is generated with Flot, a pure
Javascript plotting library for jQuery. The date-time x-values
are unix timestamps (micro seconds), the y-axis are the number
of simultaneous (concurrent) users.
File A:
A list of unixtimestamps followed with the number of
concurrent users and number of busies:
timestamp;concurrent users;busies
1291171080;1;0
1291171140;1;0
1291171200;1;1
1291171260;1;0
1291171320;1;0
1291171380;1;0
timestamp: the number of seconds elapsedsince
midnight Coordinated Universal Time (UTC) of Thursday, January
1, 1970
1291171080 = Wed, 01 Dec 2010
02:38:00 GMT
http://www.epochconverter.com/
concurrent users: number of concurrent users
busies: 'access denied' number of simultaneous
users exceeded limit
File B:
Is a file like the original file. The main differences are the
start and end time in unix timestamp. Those timestamps are
easier to work with especially in combination with file
A.
YYYY-MM-DD;hh:mm;minutes;unix timestamp start;unix timestamp
end;ip number;alias;name;e-mail;host;minutes2
In the above minutes is twice listed. The first one unaltered,
minutes and busies. The second altered, the busies are replaced
by 0.
Do not open the IPDetail csv with excel or other spreadsheet
as this may change the content and will cause an error while
importing it.
De site over Cajas, een prachtig natuurpark in Ecuador:
www.parque-nacional-cajas.org
www.uva-view.nl