1
$\begingroup$

I have a data daily rainfall from 104 stations. The problem is what are the suitable packages to solve my missing value by using R. Most of the program are suitable for multivariate data. Im prefer inverse distance weighting method. Or any other suggestion are most welcome. my data is look like this Time MM 1/Jan/1990 2.0 2/Jan/1990 NA 3/Jan/1990 11.0 and so on

all stations have missing value less than 10%

$\endgroup$
3
  • $\begingroup$ Are trying to calculate/interpolate the missing values spatially (e.g. how rain has fallen in the areas between the 104 stations?) The reason I ask is because you mention inverse distance weighting... $\endgroup$
    – user918967
    Dec 7, 2016 at 5:55
  • $\begingroup$ Can you teach me how to solve by using R. Actually im trying to compare the result (smallest RMSE) between inverse distance weighting(IDW) and K-nearest neighbour (KNN) and also multiple imputation (MI)..or can i contact you personally..sorry if to many question $\endgroup$
    – Rosbert
    Dec 7, 2016 at 6:06
  • $\begingroup$ yes im trying calculate spatially :) $\endgroup$
    – Rosbert
    Dec 7, 2016 at 6:08

2 Answers 2

1
$\begingroup$

Since you want to spatially interpolate (and IDW in particular), I would use the R package gstat see here: gstat package which also has some nice tutorials.

Here is a easy tutorial for you to read through (it uses precipitation and goes through a couple of different spatial interpolation algorithms) that should point you in the right direction Interpolation in R

$\endgroup$
1
  • $\begingroup$ just remember to vote for the answer if it works for you! $\endgroup$
    – user918967
    Dec 8, 2016 at 3:46
0
$\begingroup$

I have used MICE package in R for missing values imputation. You can take a reference from the link https://www.r-bloggers.com/imputing-missing-data-with-r-mice-package/.

$\endgroup$
1
  • $\begingroup$ Thank you so much.. will try it ..really appreciate your comment $\endgroup$
    – Rosbert
    Dec 7, 2016 at 9:20

Not the answer you're looking for? Browse other questions tagged or ask your own question.