Convertig Matlab code to R code

[florenciab]

Hello again,

I need help to convert a Matlab script to R script. I have already converted several things, but some functions or codes I can't understand them and I have searched without finding equivalences in R.
The whole script is based on a netCDF file with sea surface temperature data (sst). I managed to open it, load latitude, longitude, time, sst.
But at the moment of loading a mask to delimit the continents and to make maps only with the oceans in color, I get stuck in the conversion.

Could somebody explain me what these commands mean and how I could translate them?
I will put the Matlab code and the one I made for R. Please let me know if there is something wrong.

Mat   lon=ncread('X164.73.83.146.280.11.37.37.nc','lon');    #no problem with this
R        lon <- ncvar_get(datos_nc, "lon")

Mat  lat=ncread('X164.73.83.146.280.11.37.37.nc','lat');    #no problem with this
R       lat <- ncvar_get(datos_nc, "lat", verbose = F)

Mat  sst=ncread('X164.73.83.146.280.11.37.37.nc','sst');    #no problem with this
R       sst <- ncvar_get(datos_nc, "sst")

Mat  sst=ncread('X164.73.83.146.280.11.37.37.nc','sst');    #no problem with this
R       sst <- ncvar_get(datos_nc, "sst")

Mat  mask=ncread('lsmask.nc','mask');    #no problem with this
R       masc <- ncvar_get(mask_nc, "mask")

Mat  ind=find(mask==0);   #here I've started with problems. 
R       masc <- image(lon,lat2,masc==0)

After this I don't really know what to do. Spending today searching but have no clue what these steps mean:

mask(ind)=NaN;
sst=permute(sst,[3 1 2]);
mask_r=repmat(mask,1,1,360);mask_r=permute(mask_r,[3 1 2]);
sst=sst.*mask_r;

Thank you very much!

[arthur.t]

Seems like the objective is to multiply two three dimensional arrays together. There are a few steps beforehand to re-arrange the dimension of mask_r so that they're consistent with sst such that they can be multiplied.

The multiplication is an element-wise product (.* in Matlab, * in R), not an inner product (* in matlab, %*% in R).

[florenciab]

Thank you Arthur. Yes, I understand the idea is to multiply two three dimensional arrays together, but I'm not getting it. I've tried several different ways.
I don't understand what this means either:

mask=ncread('lsmask.nc','mask');
ind=find(mask==0)
mask(ind)=NaN

I think he wants to indicate that all the continents (mask) are left blank so that only the ocean remains in color. But I don't know how to do it in R.

I would also like to know if the function conversions (permute and repmat) I made are correct or not:

M sst=permute(sst,[3 1 2])
R sst <- aperm(sst, c(3,1,2))

M mask_r=repmat(mask,1,1,360)
R mask_r <- array(mask, c(360, 180, 360))

Thank you very much!

[arthur.t]

ind=find(mask==0)
mask(ind)=NaN

I think Matlab find is similar to R which.

These two lines of code should replace 0 values with NaN.

Perhaps in R,

ind <- which(mask == 0)
mask[ind] <- NA

[arthur.t]

On the subject of the permute vs. repmat. I think that will take come careful work. Will be tricky without access to Matlab.