Fix for Issue #120

[Max-Bladen]

Replaced the variable df with contrib.df. As plotLoadings() iterates over each block, rather than setting df as the loading values, append these loading values to the contrib.df variable. Set the name of that contrib.df list item to that of the block and return contrib.df.

[Max-Bladen]

Upon further inspection, this issue is not exclusively related to block.splsda objects. Running plotLoadings() on any multi-omics object (includes (s)pls, mint.splsda and interestingly splsda). Checks are failing due to the fact that they are seeking for a "data.frame". If we are to include all blocks' loadings values within a single output, we need a list of dataframes (as each dataframe has a different number of features, meaning it cannot be in a single dataframe).

Hence, the latest commit adjusts the values within test-plotLoadings.R such that they test pl_res against "list" rather than "data.frame" (lines: 13, 26, 45, 58)

[Max-Bladen]

While working on PR #205, I have reached the test-plotLoadings.R document. Due to the crucial changes that this PR makes to the return values of the plotLoadings() function, I will refrain from adjusting the test cases for this function until this PR has been merged. Otherwise, I would have to go back and redo these test cases in the future. When you get the chance, please review this PR @aljabadi