fixed a few bugs

      corr$.group <- c(x$Group, x$Group)

    }

  }

  mat <- matrix(nrow = length(row_names),

                ncol = length(col_names),

                dimnames = list(row_names, col_names))

  corr$.rownames <- as.integer(factor(corr$.rownames, levels = row_names))

  corr$.colnames <- as.integer(factor(corr$.colnames, levels = col_names))

  if (isTRUE(grouped)) {

    out <- lapply(split(corr, corr$.group), function(.corr) {

      r <- mat

      p <- mat

      r[.corr$.rownames, .corr$.colnames] <- .corr$r

      p[.corr$.rownames, .corr$.colnames] <- .corr$p

      r <- df_to_matrix(.corr, "r", row_id = ".rownames", col_id = ".colnames",

                        row_names = row_names, col_names = col_names)

      p <- df_to_matrix(.corr, "p", row_id = ".rownames", col_id = ".colnames",

                        row_names = row_names, col_names = col_names)

      as_correlate(x = r, p = p, is_corr = TRUE)

    })

    class(out) <- "grouped_correlate"

  } else {

    r <- mat

    p <- mat

    r[corr$.rownames, corr$.colnames] <- corr$r

    p[corr$.rownames, corr$.colnames] <- corr$p

    r <- df_to_matrix(corr, "r", row_id = ".rownames", col_id = ".colnames",

                      row_names = row_names, col_names = col_names)

    p <- df_to_matrix(corr, "p", row_id = ".rownames", col_id = ".colnames",

                      row_names = row_names, col_names = col_names)

    out <- as_correlate(x = r, p = p, is_corr = TRUE)

  }

  out