如何使用R将PDF中的表格提取到可用的tibble中

Question

我正在尝试使用R从.pdf文件中提取表。我尝试使用tabulizer包，它可以将表提取到一个很大的列表中。我想更进一步，清理这些表(它们都是不同的)并将它们放入tibble (或data.frame)中。

#incase you don't have the tabulizer package, the below is needed
install.packages("rJava")
library(rJava) # load and attach 'rJava' now
install.packages("devtools")
devtools::install_github("ropensci/tabulizer", args="--no-multiarch")
library(tabulizer)

#set path to file
file <- "https://www.sdccu.com/CURates/HomeLoanRates.pdf"

#extract tables
mortgagerates <- extract_tables(file, encoding = "UTF-8")

#first table from the third page
mortgagerates[[7]]

这是最后一行代码的输出：

> mortgagerates[[7]]
  [,1]                                                                                                                  
 [1,] "ADJUSTABLE RATE MORTGAGES: JUMBO LOANS $453,101 TO $1,500,000 
(Purchase or Refinance)"                               
 [2,] "Available for all counties:"                                                                                         
 [3,] " Purchases or refinances up to 95% LTV with a maximum loan amount of 
$679,650.  Cash-out refinances up to 70% LTV."
 [4,] ""                                                                                                                    
 [5,] " Purchases or refinances up to 80% LTV with a maximum loan amount of 
$1,500,000."                                   
 [6,] "Annual Percentage Loans Amortized Over 30 Years. Rate Rate (APR) 
Points Per $1,000 Borrowed Estimated Payment"       
 [7,] "5/1 CMT 3.500% 4.394% 0.000 $4.49"                                                                                   
 [8,] "7/1 CMT 3.750% 4.358% 0.000 $4.63"                                                                                   
 [9,] "3.500% 4.322% 1.000 $4.49" 

最好的方法是将它整理成一个类似于实际pdf文档的tibble？下表中我想要的图像：

​

​

以下是来自dput(mortgagerates[7])的更新

> file
  [,1]                                                                                                                  
 [1,] "ADJUSTABLE RATE MORTGAGES: JUMBO LOANS $453,101 TO $1,500,000 
(Purchase or Refinance)"                               
 [2,] "Available for all counties:"                                                                                         
 [3,] " Purchases or refinances up to 95% LTV with a maximum loan amount of 
 $679,650.  Cash-out refinances up to 70% LTV."
 [4,] ""                                                                                                                    
 [5,] " Purchases or refinances up to 80% LTV with a maximum loan amount of 
 $1,500,000."                                   
 [6,] "Annual Percentage Loans Amortized Over 30 Years. Rate Rate (APR) 
Points 
Per $1,000 Borrowed Estimated Payment"       
 [7,] "5/1 CMT 3.500% 4.394% 0.000 $4.49"                                                                                   
 [8,] "7/1 CMT 3.750% 4.358% 0.000 $4.63"                                                                                   
 [9,] "3.500% 4.322% 1.000 $4.49" 

Answer 1

此文件中表格的布局太复杂，如果没有更多的输入就无法自动提取它们。使用tabulizer解决这个问题的方法是提供包含表的区域。对于这个特定的表，您可以执行如下操作：

file <- "https://www.sdccu.com/CURates/HomeLoanRates.pdf"
area <- locate_areas(file, pages = 3)
area
[[1]]
      top      left    bottom     right 
442.20975  30.50972 549.83752 592.01857
mortgagerates <- extract_tables(file, pages = 3, area = area, guess = FALSE)

这提供了：

 > as.data.frame(mortgagerates[[1]])
                                                     V1         V2 V3     V4                                    V5
1 Annual Percentage Loans Amortized Over 30 Years. Rate Rate (APR)    Points Estimated Payment Per $1,000 Borrowed
2                                        5/1 CMT 3.625%     4.439%     0.000                                 $4.56
3                                        7/1 CMT 3.875%     4.417%     0.000                                 $4.70
4                                                3.625%     4.381%     1.000                                 $4.56