Errors with data.table for rMinRVar and rMPVar

[fcocquemas]

Hello,

Thanks for your hard work on the package! I've been trying to evaluate it for integrated variance estimations. I ran into a couple of errors with a data.table input, with rMinRVar and rMPVar but not with the other estimators.

library(data.table) # 1.14.8
library(highfrequency) # 1.0.0

df <- structure(list(
  DT = structure(
    c(1594128600, 1594128660, 1594128720, 
      1594128780, 1594128840, 1594128900, 1594128960, 1594129020, 1594129080, 
      1594129140, 1594129200, 1594129260, 1594129320, 1594129380, 1594129440, 
      1594129500, 1594129560, 1594129620, 1594129680, 1594129740), 
    class = c("POSIXct", "POSIXt"), tzone = "UTC"), 
  RET = c(0, 0.0003862381, -0.0016121864, -0.0009737015, 0.002679348, 
          0.0021805763, 0.00092983246, -0.00041151047, -0.0017676353, 
          -0.00049209595, -0.0022382736, -0.0020422935, 0.003068924, 
          -0.00036001205, 0.00045347214, 0.0008120537, 0.00073194504, 
          -0.0011444092, 0.0015034676, 0.00050497055)
), 
row.names = c(NA, -20L), class = c("data.table", "data.frame")                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                           
)

# Realized Variance (rRVar) works
rRVar(df, alignBy="minutes", alignPeriod=1)
# Out:
#            DT         RVar
# 1: 2020-07-07 4.398549e-05

# Median realized variance (rMedRVar) works
rMedRVar(df, alignBy="minutes", alignPeriod=1)
# Out:
#            DT     rMedRVar
# 1: 2020-07-07 5.658462e-05

# Minimum realized variance (rMinRVar) fails
rMinRVar(df, alignBy="minutes", alignPeriod=1)
# Out:
# Error: Not a matrix.

# Realized quadpower variance (rQPVar) works
rQPVar(df, alignBy="minutes", alignPeriod=1)
# Out:
#            DT          RET
# 1: 2020-07-07 2.342847e-09

# Realized multipower variance (rMPVar) fails
rMPVar(df, alignBy="minutes", alignPeriod=1)
# Out:
# Error in setnames(x, value) : 
#  Can't assign 2 names to a 1 column data.table

# Realized semivariance (rSVar) works
rSVar(df, alignBy="minutes", alignPeriod=1)
# Out:
# $`2020-07-07`
# $`2020-07-07`$rSVardownside
# [1] 1.770339e-05
# 
# $`2020-07-07`$rSVarupside
# [1] 2.62821e-05

Am I doing something wrong?

Converting to xts seems to solve the issue, but I'd rather not, if I can help it.

Thank you so much!

Best, François

[onnokleen]

Thanks for spotting this error. I found the bug and it will be fixed in the next release.

[onnokleen]

@kboudt Can you close this issue as the new version is on CRAN?

[fcocquemas]

Thank you for the fix!