Monday, May 08, 2006

The virtues of Good Database Management

This post could also be titled, the sins of not keeping on top of crap and not having a good game plan from the beginning.

After returning from Harvard I decided I would spend the day uploading all my new identifications and new pictures to my various databases. At the moment I have my Mantis database, which is supposedly the master folder of all information related to my project. This stores all my specimen identifications, collection information, location information, images, etc. I have copies of this (just in case) on the school server, a portable hard drive, my home computer, and an old lab computer. I also have my webpage which contains a species list and images, which I am trying to keep up to date. When I started downloading my new pictures I realized that I had several copies of images in different folders, and that I had been working from different folders and updating the pictures randomly from the different folders so that now I am completely at a loss as to which are the most recent and up to date photos. So, lesson 1: make sure you know where your master folder is and keep on top of that. And make sure when you are making copies of your database that you are not accidentally overwriting more recent pictures.

Secondly, I greatly regret not taking pictures of the labels of my specimens. I have several photos of ants which are identified in the title as a species I know now that I do not have. But since I don't know what specific specimen the picture is from, I am unsure as to what species it is, I will either have to redo my photos or try to re-identify them from the photos. More wasted time. So, lesson 2: take photos of labels, not just ants.

Lesson 3: pick a format for naming files and be consistent with that format. This has been a huge problem for me. As an example. I have files named OdontomachusyucatecusH1.25x, Odontomachus yucatecusH1, Odontomachus_yucatecus H1.25, etc. And they're all copies of the same picture. It's worse when you get into new species or morphospecies or queens.

Lesson 4: don't take a picture of a specimen until you've identified it. This kind of goes along with lesson 2, I suppose, since this would be less of a problem if I had taken pictures of labels from the beginning. But I discovered I have a whole bunch of images with names like Crematogaster sp 1 or Azteca sp 4, etc. At this point those photos are just useless to me because I have no idea what specimen they belong to, what species they are, nothing. It was just wasted time on my part.

Lesson 5: Keep on top of your database. For awhile I was going to Harvard very regularly and taking lots of photos and doing lots of ID's, but not updating my database at home. I just figured I would get to it later. Bad idea. It is a much better idea to take a day taking pictures, and then the next day getting all those pictures into good shape and uploading them then it would be to take pictures for several days or weeks and then try to upload everything later. You will forget things. You will make mistakes. You will realize that you forgot to put a ruler in all your photos and if only you had realized that after the first day and not after a week of taking photos! You will realize you forgot to get certain information that you will be able to enter because it was only the day before. Trust me. This is good advice. At the moment I have photos with species names on them I know I don't have. I have multiple photos of obviously different species with the same name on them. I have photos that I am basically going to have to throw away because they are useless to me. I have multiple multiple copies of files with many different names that are the same image.

Right now I am in the process of trying to get a handle on the mess that is my database so I can feel confident moving forward. I am also going to research the little button on my Windows toolbar that says "synchronize." That seems useful. I will, as always, keep you updated.

No comments:

Post a Comment