Wsrtjtyk

I have really been struggling with using R to analyze financial data. I am new to programming in general, really, except very accustomed to doing work in Excel. Consequently, I have spent a lot of time (probably too much time) formatting my CSV file, just so I could minimize the hassle when working in R, but this hasn't worked.

Here is my code for PCA analysis. I have only gotten it to work when I have used smaller data files with no N/As or blanks, but I need to know how to handle these in R.

returns <- read.csv("PCA Data File.csv", skip = 1, header = T)

#standardize the variables

returns.pca <- prcomp(returns[2:ncol(returns)], scale = TRUE)

The result is:

Error in svd(x, nu = 0) : infinite or missing values in 'x'

Many questions arise from this, the first being how do you resolve this? Second, how do I explore my data to make sure missing values are properly addressed or replaced? Is it the fact that my data is a data.frame and not matrix that is causing the issue?

I am not sure how to attach the CSV file, but here are the first few rows from the file (there are 241 rows):

Date    Returns Var1    Var2    Var3    Var4    Var5    Var6    Var7    Var8    Var9    Var10   Var11   Var12   Var13   Var14   Var15   Var16   Var17   Var18   Var19   Var20   Var21   Var22   Var23   Var24   Var25   Var26   Var27   Var28   Var29   Var30   Var31   Var32   Var33   Var34   Var35   Var36   Var37   Var38   Var39   Var40   Var41   Var42   Var43   Var44   Var45   Var46   Var47   Var48   Var49   Var50   Var51   Var52   Var53   Var54   Var55   Var56   Var57   Var58   Var59   Var60   Var61

6/30/2014   0.48    18.12   9.44    107.43  19.53   1.92    11.54   0.99    3.33    98.83   0.44    2.59    3.42    105.15  308.59  80.44   1.36    0.94    102.07  1.69    331.47  53656.02    21897.39    11022.87    23144.90    15131.80    0.59    2.70    1.35    0.58    0.33    0.25    103.38  1.67    2.59    3.42    1.75    0.10    1.09    2.00    -0.11   1.24    2.08        0.22            138780.00                                                       

5/31/2014   1.52    17.63   9.44    107.18  14.36   1.96    12.48   1.01    3.49    98.60   0.37    2.55    3.39    101.79  306.79  79.96   1.37    0.93    101.84  1.68    324.69  53122.21    21159.31    10558.07    22584.93    14343.14    0.59    2.62    1.40    0.52    0.41    0.11    103.39  1.58    2.55    3.39    1.81    0.09    1.11    1.96    -0.07   1.15    2.29        0.47    3.50    1.49    138492.00   171.04  11302.80    4322654.00  55.40   -44.39  441.59  1000.70 117.44  11.60           6.50    1.50    0.50

4/30/2014   1.07    17.40   9.45    107.11  22.93   1.96    14.20   1.02    3.49    98.24   0.40    2.69    3.52    102.03  308.63  79.85   1.38    0.93    102.51  1.67    323.24  51470.08    21660.07    10399.85    22598.44    14475.33    0.61    2.67    1.53    0.53    0.47    0.06    103.47  1.69    2.69    3.52    1.82    0.09    1.49    2.08    0.02    1.16    2.04    -4.63   0.04    3.50    1.42    138268.00   171.58  11227.50    4296049.00  54.90   -47.04  425.02  204.90  117.57  11.60       27.30   6.60    1.80    1.40

3/31/2014   0.50    17.51   9.51    106.40  25.98   1.95    14.84   1.09    3.65    98.40   0.38    2.72    3.62    100.51  303.49  79.87   1.38    0.91    102.36  1.66    316.98  47046.98    20839.70    10097.38    21980.77    14694.83    0.61    2.72    1.59    0.52    0.48    0.04    103.44  1.63    2.72    3.62    1.99    0.08    1.73    2.10    0.00    1.13    2.02        0.91    3.30    1.20    137964.00   171.47  11169.00    4226971.00  53.70   -44.18  452.77  608.80  117.39  11.70   15.10   27.30   6.80    1.60    0.20

2/28/2014   1.76    17.10   9.52    106.27  25.35   1.96    15.47   1.13    3.88    98.46   0.31    2.70    3.66    100.68  294.91  80.44   1.37    0.90    102.12  1.66    315.92  47367.89    20039.38    10048.23    22188.31    14617.57    0.60    2.74    1.66    0.44    0.44    0.01    103.45  1.50    2.69    3.66    2.16    0.07    1.82    2.10    -0.05   1.04    1.87        0.91    3.10    1.08    137761.00   169.34  11133.50    4159972.00  53.20   -42.59  383.36  -48.40  116.28  11.70       27.30   6.90    1.70    1.70

It looks like your data has problems with missing values for some of the dates so you have to do some data cleanup. The code below is an example of how you might do this for the rows you provided. Only two dates seem to be complete so continuing on to the PCA analysis didn't make much sense.

I've loaded you input data from above into the variable xx.

 xx <- sub("n"," ",xx)            #  delete n in data

 xy <- unlist(strsplit(xx,split=" "))      #  change string to character vector

 start_of_new_date <- grep("[0-9]/[0-9]{2}/2014",xy)      # find start of new dates in data

 diff(start_of_new_date)                  # notice that the number of values between dates are not all 62 so some lines are missing values

 ar <- matrix(c(c("Date", xy[1:61]), xy[168:291]), nrow=3,byrow=TRUE )       # convert only complete dates, March and April, to matrix

 df <- data.frame(Date=ar[2:3,1], ar[2:3,2:62], stringsAsFactors=FALSE)      #   convert dates and data to data frame

 colnames(df) <- c("Date",ar[1,2:62])                          # make var strings column names in data frame

 df[,2:62] <- sapply(df[,2:62], as.numeric)                  # convert data columns from character to numeric

 dfs <- scale(df[,2:62])         # example only; running scale on two row data columns is meaningless since all will scale to same values

Possible duplicate of Error in svd(x, nu = 0) : 0 extent dimensions

Negative infinity values can be replaced after a log transform as below.

log_features <- log(data_matrix[,1:8])

log_features[is.infinite(log_features)] <- -99999