[Geysers] Time code

[Lynn Stephens]

I was going to stay out of this discussion but now believe Will's use of my name in his post requires at least some response from me. I am opposed to adding a time or new code or "?" or "A" or assigning a time to entries imported from the electronic versions of the OFVC logbook into geysertimes.org.  The observer entered into the logbook the information the observer wanted to report.  I don't believe the information should be changed. I support with comments made by Jeff Cross and issues raised by MA Bellingham. From Will's comments, it appears some aspects of the OFVC logbooks don't fit into the geysertimes.org framework.  Perhaps the OFVC logbooks, especially those prior to the late 2000's shouldn't be merged into geysertimes.org.   Forcing the OFVC data to fit the geysertimes.org framework does not make sense, especially when the OFVC data is the majority, if not the the only, data available to input into geysertimes.org for years prior to the late 2000's.  (Maybe sometime in the future researchers will discover that some gazers who never entered data into either the handwritten logbooks or geysertimes.org have contributed their logbooks to a library thereby uncovering additional geyser data.  But I'm willing to bet those personal logbooks aren't going to match the geysertimes.org framework either.)
Will wrote "Now comes the questions about me applying this code to the old logs during  the transfer project but first I think it will help if I explain the current  processes and practices that Lynn has helped me put in place."   The only contribution I have made to the "processes and practices" Will has put in place has been to answer questions about specific entries in the electronic versions of the Old Faithful Visitor Center logbook.  The only reason I became involved in Will's efforts was to try to ensure that information did not get changed as it went from data base to the other, with the exception of situations where the OFVC logboook used an obsolete name or an "UNNG" that has a more recently applied name used by geysertimes.org.  I have been contacted at times by Will or his helpers asking for clarification of entries in the electronic versions of the OFVC logbook.   Most of those questions have involved obsolete geyser names, historical activity that occurred in the 1970s with which people were not familiar, etc.   Many of the questions have provided "teachable moments" in which I can provide a context within which the geyser activity occurred, tools used to collect the data, etc., although sometimes I simply have to say, "I don't know because by the time we started transcribing the logbooks, the people who could have answered the question were no longer around, so we entered just what was there."   Will made comments about difficulties he and his helpers seem to be having importing data from the OFVC logbooks into geysertimes.org and/or using data imported into geysertimes.org in Excel spreadsheets.   I am not going to address his comments other than to say there were reasons each database was set up the way it was.  What one of us sees as a "solution", the other sees as a "problem" and vice versa. Lynn Stephens  From: wolveslax65 at comcast.net
To: geysers at lists.wallawalla.edu
Date: Tue, 15 Jan 2013 10:33:15 -0700
Subject: [Geysers] Time code





Gazers,
 
The new code is necessary in my opinion because “0000?” can mean to things 
when only looking at the top level (in this case the top level is the date, time 
and codes only). By adding the “A” it will allow us once the search function is 
revamped in Geysertimes to easily select only the exact time eruptions and so 
forth in our data pulls. Also this addition would be added to the entry form 
just as another button in the row with the other codes and everyone will still 
be encouraged to put comments about the observation as we do now.
 
So lets take an example, lets say I did a data pull for a geyser and it had 
2000 entries and 200+ of them had “0000?”. I doubt many of us would want to read 
200+ entries to figure out whether it was an exact time or an overnight time. 
Then if it was an overnight time, what was the time range or the middle of the 
time range so I can manually go through the data I just pulled and change the 
0000? to the correct time frame.
 
Also by adding the “A” Geysertimes could then be programed to add a “~” to 
the interval number based on a simple has “A” or doesn’t have “A” algorithm. 
This is because the system will never be able to read and comprehend our 
comments (unless there is some easy magical way that I do not know about).
 
So in the end the final reason I want to add this is that “0000?” can mean 
two things at the top level, where as “0000A” can only mean one.
 
 
Now comes the questions about me applying this code to the old logs during 
the transfer project but first I think it will help if I explain the current 
processes and practices that Lynn has helped me put in place.
 
So first the process.
 
First I take the text file from the GOSA site and put it into Excel. This 
is the simple part.
Next comes the long part. When the text file is imported into Excel thing 
don’t always go into the right places. For example, two eruptions end up on the 
same line, data is not in the right column, etc.. . So the big task is moving 
everything around so it is in the respective column.
During this process I frequently find some weird entries. (Take a look 
sometime they are pretty confusing or funny) Whenever I hit one of these I flag 
it and send the entry to Lynn to get conformation or clarification on the entry. 
Sometimes a entry comes up where it is clear that the time was written in 12 
hour format and not in 24 hour. So I always check these with Lynn and if she 
agrees then it is changed BUT the original text is always listed with a note 
about the change the same way Lynn did when typing up the logs by using the 
“[comment]” notation.
Then comes the easy final step, the actual transfer. Once the log is 
perfect, the Excel file is saved as a csv and then uploaded to 
Geysertimes.
 
During this process it is made the top priority not to change entries. I 
also try and make that clear to my helpers as well. Almost nothing is left out 
currently  (only thing is some VR entries that have a non VR time also as 
making sure the VR entry is placed as the secondary is not possible right now.) 
this includes comments, UNNGs, notes, new geysers, etc..
 
 
Now for adding the code.
 
The reason for adding the “A” in the logs that will be transferred is the 
same as it will be for the future entries. That reason being clarity. In 
databases it is best to keep notation the same throughout the entire data. 
Currently with the notes not being visible in the data retrieve this sort of 
forces me to change the entry from a note to an eruption that is “0000?”. Again 
that can mean two things.
 
 
Now the question about why we need this now. My reason is simple, I don’t 
want to have to go through and clean the transfers of the logs. If I can make 
this clarification with the code “A” this first time through it will be much 
easier than having to go through everything again. Relooking through about 30-40 
years of logs with a minimum of 5000-6000 entries a year doesn’t sound too fun 
to me.
 
Now about filling the gaps. This code will not fill any data gaps, it will 
only clarify the current data that we already have. The only way we fill the 
data gaps is by people submitting missing data. If you are interested in doing 
that please contact me.
 
I hope this helps,
 
Will Boekel
 
 

_______________________________________________
Geysers mailing list
Geysers at lists.wallawalla.edu
 

 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: </geyser-list/attachments/20130116/4411facb/attachment-0001.html>