davidatlarge / rec_r Goto Github PK

add error handling were useful. E.g. in:
reading args from table

check if there are any objects created as output by helper functions that are not used anywhere.

does z (depth) have to be positive? if so, should rec() allow input of negative z or should a warning be printed if z is negative?

At the moment providing the input of test_case_2, pre-processing and calling the function calculate_con_rates_lin_sys_tichonov_mean_rate_2() does produce a result, but it is obviously wrong.

I found that the REC_R package cannot be installed no matter using commend or install via the .zip file.
Is there any method to install the package?
install.packages("https://github.com/markolipka/REC_R")
Warning in install.packages :
package ‘https://github.com/markolipka/REC_R’ is not available for this version of R.

calculate_diff_operator_matrix_aequi_dist_grid_variable_coeff() could be tested with output of original Matlab-code available for certain edge-cases.

Output of test_case-scenario would be helpful to at least check if the function works in the simplest case as desired by the original author.

This function seems to be a very strange workaround to read variables from a parent environment into calculate_diff_operator_matrix_aequi_dist_grid_variable_coeff().
This should be rather done by passing the needed values as arguments to the function.

make_column_vector() and make_row_vector() may be unnecessary, as with the test_case_2_data the result is the same without converting vectors to one-dimensional matrices

Would be helpful to have the original output of all intermediate steps for a simple test case (e.g. test_case_2) so that each REC_R function can be tested for at least this test case.

Originally posted by @markolipka in #4 (comment)

replace the use of pracma:: and matlab:: with base R functions