在 R 中：按日期从数据框中提取特定列，并将基本列保留在开头

Question

I have several big data frames with time series data for specific coordinates from 2007-2019. 8113 rows and 301 columns. Each year is divided in 16 days time steps, which results in 23 values per year per coordinate. It looks like this:

X   longitude   latitude   label     2007-01-07   2007-01-23 ... 2019-12-10   2019-12-26  
1   -56.58652   -30.87850  cropland  0.08367160   0.07883158     0.07414120   0.08120061
2   -56.58458   -30.88260  cropland  0.07888613   0.07438400     0.07831833   0.07352642
3   -56.58429   -30.87860  cropland  0.08331446   0.07837244     0.07169452   0.07229450

I would like to subset the data frame into years, keeping the first four columns the same for each subset. Then afterwards save them all together in a list of dataframes. So the output I am looking for should be like this:

X   longitude   latitude   label     2007-01-07   2007-01-23   ...
1   -56.58652   -30.87850  cropland  0.08367160   0.07883158   ...  
2   -56.58458   -30.88260  cropland  0.07888613   0.07438400   ...  
3   -56.58429   -30.87860  cropland  0.08331446   0.07837244   ...

X   longitude   latitude   label     2008-01-10   2008-01-26   ...
1   -56.58652   -30.87850  cropland  0.08367160   0.07883158   ...  
2   -56.58458   -30.88260  cropland  0.07888613   0.07438400   ...  
3   -56.58429   -30.87860  cropland  0.08331446   0.07837244   ...

...

X   longitude   latitude   label     2019-01-12   2019-01-28   ...
1   -56.58652   -30.87850  cropland  0.08367160   0.07883158   ...  
2   -56.58458   -30.88260  cropland  0.07888613   0.07438400   ...  
3   -56.58429   -30.87860  cropland  0.08331446   0.07837244   ...

I need to do that for 8 dataframes like the example above. I know, this should be quite basic, but I'm also quite new to R and programming in general. So I'm thankful for any hint on that one! Cheers!

Answer 1

You can use split.default to split data based on year and with lapply cbind the first four columns to each list.

result <- lapply(split.default(df[-(1:4)], 
                 format(as.Date(names(df)[-(1:4)], 'X%Y.%m.%d'), '%Y')), 
                 function(x) cbind(df[1:4], x))

R tries to discourage column names starting with numbers so if you read the data with default options it will change column name from 2007-01-07 to X2007.01.07 so keeping that in mind I have used 'X%Y.%m.%d' in as.Date . If you have somehow managed to read column names as you have shown ie 2007-01-07 use %Y-%m-%d in as.Date .