This was the case in my course which is markable , for it have been my first time to accomplish a case.
- Dataset : UFO-sighting records throughout the world
- Purpose : Visualize for the following analysis
- "ufo_sightings.pdf" -> case purpose
- "Number of UFO sightings by Month−Year and U.S. State (1990−2010).pdf" -> my final result
- "Number of UFO sightings by Month−Year and U.S. State (1980−2001).pdf" -> my result of first stage(first 10000 obs.)
- "ufo(first 10000 obs.).R" -> R code for the first 10000 obs.
- "ufo(full data).R" -> R code for the full data
-
The codes are somewhat not concise , because I wanted to preserve my original train of thought , in contrast to improvements which would be made afterwards.
-
Comments amid codes are, partly in Chinese ,partly in English. Its embarrassed ,due to there being some bugs in Input Method.
When teacher assigned this case ,I took it serious. And it cost about 3 days to do this.
I began the case with very limited knowledge on R . To achive every periodical targets , I refered to tutorials, books ,Internet, as well as read source codes of R functions.
For instance, to come out with the ways of reading every single character of a string , I went over source code of function "as.Date.character" thinking of now that "as.Date.character" is able to identify every character of a string , there would be some codes inspiring me.
Then ,function "strptime" was found out . Furthermore ,the class of function begun with "str-" emerged .At last , I picked "strsplit" to realize my identification ,ideally.
From the aspect of complexity , I have realized the case's purpose on first 500 ,first 10000 ,full dataset ,successively.
By doing this case ,I have had a more profound understanding of explory learning and got a sense of achivement.