R matrix - how to convert data from data frame columns into 1/0 matrix-CodePudding

I'm preparing a master's degree project and stuck with basic data manipulation. I'm importing several data from the Prestashop database to R, one of those is a data frame with carts IDs and products included in it (see below).

What I want to do is to create a matrix that will reflect the same data but in the easiest way as a matrix, here's a draft of the most desirable look:

Any hints on how the code should look? Thank you in advance for any help!

EDIT:

Code sample (dataframe):

x <- data.frame (order_id  = c("12", "13","13","13","14","14","15","16"),
                   product_id = c("123","123","378","367","832","900",NA,"378"))

SOLUTION:

xtabs is good, but when it comes to NA values it skips the line in the results. There's an option to force addNA=TRUE, but it adds the NA 'column' and counts the NA as 1 (see below)

y <- xtabs(formula = ~., data = x)

Output - example 1 (addNA=FALSE):

        product_id
order_id 123 367 378 832 900
      12   1   0   0   0   0
      13   1   1   1   0   0
      14   0   0   0   1   1
      16   0   0   1   0   0

Output - example 2 (addNA=TRUE):

    product_id
order_id 123 367 378 832 900 <NA>
      12   1   0   0   0   0    0
      13   1   1   1   0   0    0
      14   0   0   0   1   1    0
      15   0   0   0   0   0    1
      16   0   0   1   0   0    0

The igraph approach seems to be more accurate.

CodePudding user response：

You are looking for creating an adjacency matrix from a bipartite network from which you have the nodes list. You can directly use the package igraph to create the adjacency matrix from the node list and simplify it.

From x:

  order_id product_id
1       12        123
2       13        123
3       13        378
4       13        367
5       14        832
6       14        900
7       15       <NA>
8       16        378

graph_from_dataframe <- igraph::graph.data.frame(x)
adjacency_matrix <- igraph::get.adjacency(graph_from_dataframe, sparse = FALSE)
# removing redundant entries
adjacency_matrix <- adj[rownames(adj) %in% x$order_id, colnames(adj) %in% x$product_id]

   123 378 367 832 900
12   1   0   0   0   0
13   1   1   1   0   0
14   0   0   0   1   1
15   0   0   0   0   0
16   0   1   0   0   0

More resources on this SO question and this RPubs blog post.