Empirical Research in Economics with R

Description

Functions, datasets, and sample codes related to the book of 'Empirical Research in Economics: Growing up with R' by Dr. Changyou Sun are included (ISBN 9780996585408). Marginal effects for binary or ordered choice models can be calculated. Static and dynamic Almost Ideal Demand System (AIDS) models can be estimated. A typical event analysis in finance can be conducted with several functions included.

Close to 100 sample programs and data sets are used in the book. They are included in the package folder on your local drive, e.g., c:/.../R-3.2.2/library/erer/doc/. Alternatively, download the tar.gz version of this package from the CRAN site to view them.

Details

Author(s)

Changyou Sun (edwinsun258@gmail.com)

Transforming Raw Data for Static AIDS Model

Description

This function transforms import values and quantities into a data format that are needed for a static AIDS model.

Usage

aiData(x, label, label.tot = "WD", prefix.value = "v",
  prefix.quant = "q", start = NULL, end = NULL, stone = TRUE, 
  dummy = NULL, season = c("none", "m", "q"), ...)

Arguments

Details

This transforms raw import data into a format needed for a static AIDS model. This separation of data prepraration from model fitting allows greater flexibility in using aiStaFit in estimating a static AIDS model. In addition, when the raw data contain zero, a small number is substituted to avoid NA when the price variable (value/quantity) is calculated. The values for the residual supplier will change when the label is different.

Value

Return a list object with two components:

Methods

One method is defined as follows:

print:

print the first several observations of the final data needed for the AIDS model.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Moschini, G. 1995. Units of measurement and the Stone index in demand system estimation. American Journal of Agricultural Economics 77(1):63-68.

Singh, K., M.M. Dey, and G. Thapa. 2011. An error corrected almost ideal demand system for crustaceans in the United States. Journal of International Food & Agribusiness Marketing 23(3):271-284.

Wan, Y., C. Sun, and D.L. Grebner. 2010. Analysis of import demand for wooden beds in the United States. Journal of Agricultural and Applied Economics 42(4):643-658.

See Also

aiStaFit; daBedRaw; daBed.

Examples

data(daBedRaw)
dumm <- list(dum1 = c(2003, 10, 2003, 10), dum2 = c(2004, 7, 2004, 7),
  dum3 = c(2005, 1, 2005, 1))
imp8 <- aiData(x = daBedRaw,
  label = c("CN", "VN", "ID", "MY", "CA", "BR", "IT"),
  label.tot = "WD", prefix.value = "v", prefix.quant = "q",
  start = c(2001, 1), end = c(2008, 12), dummy = dumm)
imp4 <- update(imp8, label = c("CN", "VN", "ID"))
imp5 <- update(imp4, label = c("CN", "VN", "ID", "MY"))
imp8; imp4; imp5
daTest <- imp8$out
colnames(daTest)[18:20] <- c("dum1", "dum2", "dum3")

data(daBed)
identical(daBed, daTest)

Diagnostic Statitics for Static or Dynamic AIDS Model

Description

Report a set of diagnostic statistics for static or dynamic AIDS models

Usage

aiDiag(x, digits = 3, ...)

Arguments

Details

Compute several diagnostic statistics for each equation in a AIDS model. Tests includes are BG, BP, RESET, and JB. See the reference paper for detail.

Value

Return a data frame object with the statistics and p values for the four tests by equation.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Wan, Y., C. Sun, and D.L. Grebner. 2010. Analysis of import demand for wooden beds in the United States. Journal of Agricultural and Applied Economics 42(4):643-658.

See Also

aiStaFit; aiDynFit.

Examples

# see the examples for 'aiDynFit'.

Fitting a Dynamic AIDS Model

Description

Estimate a dynamic AIDS model for a system.

Usage

aiDynFit(w, dum.dif = FALSE, AR1 = FALSE,
  rho.sel = c("all", "mean"), ...)

Arguments

Details

This estimates a dynamic AIDS model. The residuals from the statis AIDS model are included. As it is programmed now, only one lag is allowed for the share variables on the right-hand side. Autocorrelation in the residuals can be corrected following the treatment in Berndt (1975).

Value

Return a list object of class "aiFit" and "aiDynFit" with the following components:

Methods

One method is defined as follows:

print:

print the first several observations of the final data.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Berndt, E.R., and N.E. Savin. 1975. Estimation and hypothesis testing in singular equation systems with autoregressive disturbances. Econometrica 43(5/6):937-957.

Wan, Y., C. Sun, and D.L. Grebner. 2010. Analysis of import demand for wooden beds in the United States. Journal of Agricultural and Applied Economics 42(4):643-658.

See Also

systemfitAR; aiStaFit; aiDiag; aiElas; summary.aiFit.

Examples

# --- Step 1: Read data
data(daExp, daBedRaw, daBed)

# --- Step 2: Hausman Test 
# 2.1 Getting started with a static AIDS model 
sh  <- c("sCN", "sVN", "sID", "sMY", "sCA", "sBR", "sIT", "sRW")
pr  <- c("lnpCN", "lnpVN", "lnpID", "lnpMY",  
         "lnpCA", "lnpBR", "lnpIT", "lnpRW")
du3 <- c("dum1","dum2","dum3")
rSta <- aiStaFit(y = daBed, share = sh, price = pr, shift = du3, 
  expen = "rte", omit = "sRW", hom = TRUE, sym = TRUE)
summary(rSta)

# The following steps should work. It can take a few minutes.

# 2.2 The final Hausman test and new data
(dg <- daExp[, "dg"])
rHau <- aiStaHau(x = rSta, instr = dg, choice = FALSE)
names(rHau); colnames(rHau$daHau); colnames(rHau$daFit); rHau
two.exp <- rHau$daFit[, c("rte", "rte.fit")] 
bsStat(two.exp, digits = 4)
plot(data.frame(two.exp)); abline(a = 0, b = 1)
daBedFit <- rHau$daFit

# --- Step 3: Static and dynamic AIDS models 
# 3.1 Diagnostics and coefficients
hSta  <- update(rSta, y = daBedFit, expen = "rte.fit")
hSta2 <- update(hSta, hom = FALSE, sym = FALSE) 
hSta3 <- update(hSta, hom = FALSE, sym = TRUE)
hSta4 <- update(hSta, hom = TRUE,  sym = FALSE)
lrtest(hSta2$est, hSta$est) 
lrtest(hSta2$est, hSta3$est)
lrtest(hSta2$est, hSta4$est)

hDyn  <- aiDynFit(hSta)
hDyn2 <- aiDynFit(hSta2); lrtest(hDyn2$est, hDyn$est)
hDyn3 <- aiDynFit(hSta3); lrtest(hDyn2$est, hDyn3$est)
hDyn4 <- aiDynFit(hSta4); lrtest(hDyn2$est, hDyn4$est)

(table.2 <- rbind(aiDiag(hSta), aiDiag(hDyn)))
(table.3 <- summary(hSta))
(table.4 <- summary(hDyn))
                          
# 3.2 Elasticity calculation
es <- aiElas(hSta); esm <- es$marsh
ed <- aiElas(hDyn); edm <- ed$marsh
esm2 <- data.frame(c(esm[1:2, 2], esm[3:4, 3], 
  esm[5:6, 4], esm[7:8, 5], esm[9:10, 6], esm[11:12, 7], 
  esm[13:14, 8], esm[15:16, 9]))
edm2 <- data.frame(c(edm[1:2, 2], edm[3:4, 3], 
  edm[5:6, 4], edm[7:8, 5], edm[9:10, 6], edm[11:12, 7], 
  edm[13:14, 8], edm[15:16, 9]))
eEM <- cbind(es$expen, esm2, ed$expen[2], edm2) 
colnames(eEM) <- c("Country", "LR.expen", "LR.Marsh", 
  "SR.expen", "SR.Marsh")
(table.5 <- eEM[-c(15:16),])
(table.6a <- es$hicks[-c(15:16), -9])
(table.6b <- ed$hicks[-c(15:16), -9])

Computing Elasticity for Static or Dynamic AIDS Models

Description

Calculate expenditure elasticity, Marshalllian price elasticity, Hicksian price elasticity, and their variances for static or dynamic AIDS Models.

Usage

aiElas(z, ...)

Arguments

Details

Calculate expenditure elasticity, Marshalllian price elasticity, and Hicksian price elasticity for static or dynamic AIDS Models. The related variance, t-ratio, p-value, and significance are also reported.

Value

Return a list object with the following components:

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Wan, Y., C. Sun, and D.L. Grebner. 2010. Analysis of import demand for wooden beds in the United States. Journal of Agricultural and Applied Economics 42(4):643-658.

See Also

aiStaFit; aiDynFit.

Examples

# see the examples for 'aiDynFit'.

Fitting a Static AIDS Model

Description

Estimate a static AIDS model for a system.

Usage

aiStaFit(y, share, price, expen, shift = NULL, omit = NULL,
  hom = TRUE, sym = TRUE, AR1 = FALSE, rho.sel = c("all", "mean"), ...)

Arguments

Details

This estimates a static AIDS model. The data supplied should be in the final format. Autocorrelation in the residuals can be corrected following the treatment in Berndt (1975).

Value

Return a list object of class "aiFit" and "aiStaFit" with the following components:

Methods

One method is defined as follows. This is the print method related to three functions: aiStaFit, aiDynFit, and aiStaHau.

print:

print the first several observations of selectec outputs.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Berndt, E.R., and N.E. Savin. 1975. Estimation and hypothesis testing in singular equation systems with autoregressive disturbances. Econometrica 43(5/6):937-957.

Wan, Y., C. Sun, and D.L. Grebner. 2010. Analysis of import demand for wooden beds in the United States. Journal of Agricultural and Applied Economics 42(4):643-658.

See Also

aiDiag; aiElas; summary.aiFit; aiDynFit; aiStaHau; systemfitAR.

Examples

# see the examples for 'aiDynFit'.

Conducting a Hausman Test on a Static AIDS Model

Description

Conduct a Hausman test on a static AIDS model and report the result of likelihood ratio test.

Usage

aiStaHau(x, instr, choice = FALSE,  
  exog = c("none", "all", "partial"), shift.new = NULL)

Arguments

Details

Conduct a Hausman test on a static AIDS model and report the result of likelihood ratio test. Note that logarithm is taken on every variable in the auxiliary regression. The dependant variable is the real total expenditure. The independant variables include the lagged value of the real total expenditure, the instrumental variable as supplied, the price variables, and some or all shift variables as included in the AIDS model.

Value

Return a data frame object with the statistics and p values for the four tests by equation.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Wan, Y., C. Sun, and D.L. Grebner. 2010. Analysis of import demand for wooden beds in the United States. Journal of Agricultural and Applied Economics 42(4):643-658.

See Also

aiStaFit; print.aiFit.

Examples

# see the examples for 'aiDynFit'.

Generating Formula for Models

Description

Generate a single formula for models like lm or a list of formula for models like systemfit.

Usage

bsFormu(name.y, name.x, intercept = TRUE, ...)

Arguments

Details

This function can generate a single formula for simple model like lm or a list of formula for systems (systemfit. Note that the right-hand side variables are the same for each dependent variable. If different, a for loop can be added by users to address that, as demonstrated by the example below.

Value

a single formula object or a list of formula objects.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

Examples

# fake data
y  <- c("y")
ym <- c("y1", "y2", "y3")
x <- c("x")
xAll <- c("x", "xx", "xxx", "xxxx")

bsFormu(name.y = y, name.x = x)
bsFormu(name.y = ym, name.x = xAll)
fm.ym <- bsFormu(name.y = ym, name.x = xAll, intercept = FALSE)
fm.ym

# If independent variables differ by equation,
# add a loop to address the differentiation.
xInd <- c("x1", "x2", "x3")
fm.ym <- list()
for (i in 1:length(ym)) {
    ny <- ym[i]
    nx <- c(xInd[i], xAll)
    fm.ym[[i]] <- bsFormu(name.y = ny, name.x = nx, intercept = FALSE)
}
fm.ym

# real data
data(daIns)
(xx <- colnames(daIns)[-c(1, 14)])
fm.ins <- bsFormu(name.y = "Y", name.x = xx, intercept = TRUE)
fm.ins
(ra <- glm(formula = fm.ins,
          family = binomial(link="logit"),
          data = daIns, x = TRUE))

Lagged Time Series

Description

Generate a set of lagged time series for time series data.

Usage

bsLag(h, lag, prefix = "", var.name, suffix = ".t_", 
    include.orig = TRUE, by.lag = FALSE, ...)

Arguments

Details

The input data can be a single time series or a set of multiple time series data. The output is a set of lagged time series with the specified lag dimension. All the series are aligned with the shortest window so the loss of observations is equal to lag. The original series (e.g., without lag but just loss of beginning observations) can be included or excluded by setting the logical value of include.org.

The name of the output data is composed of four parts: prefix, var.name, suffix, and an index number of lag. Users can control the first three parts only because the lag number is added automatically. prefix and suffix can be fixed for all the output series. var.name provides some flexibility when bsLag is used within a function and the variable name is unknown a priori.

The column of the output can be ordered either by the variable name (e.g., diff.GA.t_0, diff.GA.t_1, diff.ND.t_0, diff.ND.t_1), or by the lag order ((e.g., diff.GA.t_0, diff.ND.t_0, diff.GA.t_1, diff.ND.t_1).

Value

Return a multiple time series object.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

Examples

# simple example
h1 <- ts(data=cbind(1:24), start=c(2001, 1), frequency=12)
h2 <- ts(data=cbind(1:24, 25:48), start=c(2001, 1), frequency=12)
h3 <- ts(data=cbind(1:4, 5:8, 9:12), start=c(2001, 1), frequency=4)
colnames(h2) <- c("aa", "bb")
colnames(h3) <- c("cc", "dd", "ee")
h1; h2; h3

bsLag(h=h1, lag=0, prefix="", suffix=".t_")
bsLag(h=h1, lag=2, prefix="price.", var.name="fl", suffix=".t_")
bsLag(h=h1, lag=2, prefix="price.", var.name="fl", suffix=".t_", by.lag=TRUE)
bsLag(h=h1, lag=23, prefix="price.", suffix=".t_", include.orig=FALSE)

bsLag(h=h2, lag=4, prefix="", suffix=".t_", include.orig = TRUE)
bsLag(h=h2, lag=4, prefix="", suffix=".t_", include.orig = FALSE)
bsLag(h=h2, lag=4, prefix="", suffix=".t_", include.orig = FALSE, by.lag=TRUE)
bsLag(h=h2, lag=0, prefix="", var.name=c("nc", "sc"), suffix=".t_")

bsLag(h=h3, lag=2, prefix="", suffix=".t_", include.orig=FALSE)
bsLag(h=h3, lag=1, prefix="", var.name=c("nd", "sd", "mi"), suffix=".lag.")
bsLag(h=h3, lag=2, prefix="NY.", suffix=".t_", by.lag=TRUE)
bsLag(h=h3, lag=3, prefix="NY.", suffix=".t_", include.orig=FALSE)

# with real data
data(daBedRaw)
small <- daBedRaw[, c("vCN", "qCN")]
(lag.small <- bsLag(h=small, lag=4))
colnames(lag.small)

resid <- residuals(lm(qCN ~ vCN, data = small))
res <- ts(resid, start=start(small), end=end(small),
    frequency=tsp(small)[3])
lag.res <- bsLag(h=res, lag=2, prefix="resid.", var.name="china")
str(lag.res)
headts(lag.res)
tailts(lag.res)

Summary of Basic Statistics

Description

Calcluate basic statistics of data.

Usage

bsStat(y, two = NULL, digits = c(2, 2), use = 'complete.obs', 
  na.rm = TRUE, ...)

Arguments

Details

Two set of summary statistics are generated. One is correlation coefficients and the other is mean, minimum, maximum, standard deviation, and number of observations. When two is unspecified and the number of variables is bigger than ten, the two sets are reported separately; otherwise, it is reported as a single data frame object.

Value

A dataframe or list of the summary statistics.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

Examples

data(daIns)
(sum.daIns <- bsStat(y=daIns, digits=c(3,2)))

Generating Pretty Statistical Tables

Description

Format statistics from regressions into pretty outputs

Usage

bsTab(w,
  need = c("1T", "1E", "2T", "2E", "3T", "3E", "4T", "4E", "5"),
  wrap.TE = c("(", "", "["),
  add.sig = c("coef", "TE"),
  percent = c(0.01, 0.05, 0.10),
  symbol = c("***", "**", "*", ""),
  digits = c(3, 3, 3, 3), ... )

Arguments

Details

Format statistics from regressions into tables that are often reported in economic journals. The column of 'Variable' in the outuput is the row names of the input data so the raw data should contain meaningful rownames. Besides the variable name column, the maximum number of output is five columns: estimate, error, t ratio, p value, and significance. wrap.TE and add.sig are only valid for column widths of 1 and 2.

Value

A dataframe of statistical results.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

Examples

# a simulated data
tes <- data.frame(est = c(4, 56, 12), err = c(0.3, 0.56, 0.789), 
  t.rat = c(2.56, 7.9, 1.2), p.val = c(0.002, 0.23, 0.061))
tes
bsTab(tes)
bsTab(w = tes, need = "2E")

# real data
data(daIns)
ra <- glm(formula = Y ~ Injury + HuntYrs + Nonres + 
                    Lspman + Lnong + Gender + Age + 
                    Race + Marital + Edu + Inc + TownPop, 
          family = binomial(link="logit"),
          data = daIns, x = TRUE)
(ca <- data.frame(summary(ra)$coefficients))

# an object of class 'glm' as input
bsTab(w = ra, add.sig = "TE")
bsTab(w = ra, wrap.TE = "[")
bsTab(w = ra, need = "5")
bsTab(w = ra, need = "4T", wrap.TE = "[")
final <- bsTab(w = ra, need = "3T", 
  percent = c(0.01, 0.05, 0.10),
  symbol = c("a", "b", "c", ""), digits = 4)
final
print(final, right = FALSE) 

# any matrix with at least four columns can be supplied
cbind(bsTab(ca), bsTab(ra)) 

Transformed Wooden Beds Import Data for Static AIDS Models

Description

This data set contains transformed values related to wooden beds imports by the United States from January 2001 to December 2008. There are 96 observations and 20 variables.

Usage

data(daBed)

Format

Monthly time series from Janurary 2001 to December 2008 with 96 observations for each of the 20 variables.

Details

This is the transformated data set for static AIDS model. The transformation detail is described in Wan et al. (2010).

Source

U.S. ITC, 2010. Interactive tariff and trade data web. US International Trade Commission (Assecced on March 1, 2010).

References

Wan, Y., C. Sun, and D.L. Grebner. 2010. Analysis of import demand for wooden beds in the United States. Journal of Agricultural and Applied Economics 42(4):643-658.

See Also

aiStaFit; daBedRaw.

Examples

data(daBed)
class(daBed); dim(daBed); colnames(daBed)
daBed

Wooden Beds Import Data

Description

This data set contains a mulitiple time series related to wooden beds imports by the United States. The time covered is January 1996 to December 2008 with 156 observations. There are 34 variables in total: 17 import values (dollars) and 17 import quantities (dollars / piece). In total, 16 countries are covered and the world total is also reported.

Usage

data(daBedRaw)

Format

Monthly time series from Janurary 1996 to December 2008 with 156 observations for each of the 34 variables.

Details

Under the Harmonized Tariff Schedule (HTS) system, the commodity of wooden beds is classified as HTS 9403.50.9040. The monthly cost-insurance-freight values in dollar and quantities in piece are reported by country from U.S. ITC (2010).

Source

U.S. ITC, 2010. Interactive tariff and trade data web. US International Trade Commission (Assecced on March 1, 2010).

References

Wan, Y., C. Sun, and D.L. Grebner. 2010. Analysis of import demand for wooden beds in the United States. Journal of Agricultural and Applied Economics 42(4):643-658.

See Also

aiStaFit; daBed.

Examples

data(daBedRaw)
class(daBedRaw); dim(daBedRaw); colnames(daBedRaw)

Daily returns of 14 public firms from 1990 to 2004

Description

This data set contains daily returns of 14 public firms, three-month treasury bill, and SP 500 Index from 1990 to 2004.

Usage

data(daEsa)

Format

A data frame object with daily returns for firms or indexes from 1994 to 2004. There are 17 columns and 3747 rows. The date is not regular because there is no trading on weekends and holidays. Therefore, the date is represented by a number, not a date.

Details

This is the transformated data set used in the study of Sun and Liao (2011).

Source

The daily returns for SP 500 and individual firms are from the database of the Center for Research in Security Prices (CRSP). The risk-free rate of return is the secondary market rate for the 3-month US Treasury bills from the Federal Reserve Bank.

References

Sun, C., and X. Liao. 2011. Effects of litigation under the Endangered Species Act on forest firm values. Journal of Forest Economics 17(4):388-398.

See Also

evReturn; evRisk.

Examples

data(daEsa) 
dim(daEsa); colnames(daEsa)
head(daEsa); tail(daEsa)
str(daEsa)

# if dates are stored as a date object in R, then it can be converted into
# numbers as the following example shows.

raw <- as.Date(c('1990-01-02', '1991-11-12')); raw; str(raw)
raw2 <- as.numeric(strftime(raw, format = "%Y%m%d")); raw2; str(raw2)

Expenditure Data for a Hausman Test in AIDS Model

Description

This data set contains seven monthly times series for expenditure from 2001 to 2008.

Usage

data(daExp)

Format

Monthly time series from Janurary 2001 to December 2008 with 96 observations for each of the seven variables.

Details

This is the data set for conducting a Hausman test in a static AIDS model, as detailed in Wan et al. (2010). The test focuses on whether the expenditure variable in a AIDS model is exogenous or not. Each of the seven expenditure data can be used as an instrumental variable in an auxiliary regression.

Source

Federal Reserve Bank of St. Louis. Economic Data - Fred. Internet site: http://stlouisfed.org (Accessed February 25, 2010).

References

Wan, Y., C. Sun, and D.L. Grebner. 2010. Analysis of import demand for wooden beds in the United States. Journal of Agricultural and Applied Economics 42(4):643-658.

Examples

data(daExp)
class(daExp); dim(daExp); colnames(daExp)
daExp

Liability Insurance Coverage for Hunters and Anglers in Mississippi

Description

This data set contains a survey result about liability insurance purchase decision by hunters and anglers in Mississippi. There are 1653 observations for 14 variables.

Usage

data(daIns)

Format

A cross sectional data with 1653 observations and 14 variables.

Details

The data set is from a telephone survey conducted in 2005 in Mississippi.

Source

Sun, C., S. Pokharel, W.D. Jones, S.C. Grado, and D.L. Grebner. 2007. Extent of recreational incidents and determinants of liability insurance coverage for hunters and anglers in Mississippi. Southern Journal of Applied Forestry 31(3):151-158.

See Also

daInsNam

Examples

data(daIns)
class(daIns); dim(daIns)
head(daIns); tail(daIns)

ra <- glm(formula = Y ~ Injury + HuntYrs + Nonres +
                    Lspman + Lnong + Gender + Age +
                    Race + Marital + Edu + Inc + TownPop,
          family = binomial(link="logit"),
          data = daIns, x = TRUE, y= TRUE)
names(ra); summary(ra)

(ins.me <- maBina(w = ra))
(ins.mt <- maTrend(q=ins.me, nam.c="Age", nam.d="Nonres"))
plot(ins.mt)

Liability Insurance Coverage for Hunters and Anglers in Mississippi: Variable Definitions

Description

This data set contains variable defintions for the a survey result about liability insurance purchase decision by hunters and anglers in Mississippi, as documented in daIns. It contains 14 observations and 2 columns.

Usage

data(daInsNam)

Format

A data frame with 2 columns and 14 rows.

Details

The data set contains the definitions of 14 variables collected from a telephone survey as detailed in daIns

Source

Sun, C., S. Pokharel, W.D. Jones, S.C. Grado, and D.L. Grebner. 2007. Extent of recreational incidents and determinants of liability insurance coverage for hunters and anglers in Mississippi. Southern Journal of Applied Forestry 31(3):151-158.

See Also

daIns

Examples

data(daInsNam)
str(daInsNam); dim(daInsNam); daInsNam

Data for statutory reforms and retention of prescribed fire liability laws on forest land by state

Description

This data set contains a cross-sectional data set for current adoption of statutory laws with regard to prescribed fire liability on forest landowners. It has 50 observations and 16 variables.

Usage

data(daLaw)

Format

A data frame object with 50 rows and 16 variables.

Details

This is the final data set used in the study of Sun (2006).

Source

See Table 2 in Sun (2006) for detail of data sources.

References

Sun, C. 2006. State statutory reforms and retention of prescribed fire liability laws on US forest land. Forest Policy and Economics 9(4):392-402.

Examples

data(daLaw)
str(daLaw); head(daLaw); names(daLaw) 

Program Effectiveness of a New Method of Teaching Economics

Description

This data set contains the evaluation results of a new program of teaching in economics. There are 32 observations for 4 variables.

Usage

data(daPe)

Format

A data frame of cross sectional data with 32 observations and 4 variables.

Details

Evaluation results on 32 students of the impact of a new teaching methods.

Source

Spector, L.C., and M. Mazzeo. 1980. Probit analysis and economic education. Journal of Economic Education 11(2):37-44.

Examples

data(daPe)
dim(daPe)
summary(daPe)
daPe

Voting records for the Healthy Forests Restoration Act in 2003 and the associated characteristics of congressmen

Description

This data set contains the voting records for the Healthy Forests Restoration Act in 2003, as used in Sun (2006). The characteristics of individual congressmen are also included. There are 537 observations and 22 variables.

Usage

data(daRoll)

Format

A data frame object with 537 rows and 22 variables. This is a cross-sectional dataset that are generating from merging several raw datasets.

Details

This is the combinded final data set used in the study of Sun (2006).

Source

See Table 1 in Sun (2006) for detail.

References

Sun, C. 2006. A roll call analysis of the Healthy Forests Restoration Act and constituent interests in fire policy. Forest Policy and Economics 9(2):126-138.

See Also

glm; maBina.

Examples

# generate four datasets used in Sun (2006)
data(daRoll)
xn <- c('RepParty', 'East', 'West', 'South', 'PopDen', 
  'PopRural', 'Edu', 'Income', 'FYland', 'Size', 
  'ContrFY', 'ContrEN', 'Sex', 'Lawyer', 'Member', 'Year', 'Chamber')

f1 <- daRoll[!is.na(daRoll$voteMay), c('voteMay', xn)]
f2 <- daRoll[!is.na(daRoll$voteNov) & daRoll$Chamber == 1, c('voteNov', xn)]
f3 <- daRoll[!is.na(daRoll$voteNov), c('voteNov', xn)]
f4 <- daRoll[!is.na(daRoll$voteNov) & daRoll$RepParty == 0, c('voteNov', xn)]
rownames(f1) <- 1:nrow(f1); rownames(f2) <- 1:nrow(f2)
rownames(f3) <- 1:nrow(f3); rownames(f4) <- 1:nrow(f4)
colnames(f1)[1] <- colnames(f2)[1] <- 'Vote'
colnames(f3)[1] <- colnames(f4)[1] <- 'Vote'
dim(f1); dim(f2); dim(f3); dim(f4)
tail(f3) 

Downloading source and PDF version of R packages

Description

This function can download three documents for each package: zip version, source version (tar.gz), and PDF manual.

Usage

download.lib(pkgs, destdir = tempdir(), mode = "wb", 
  pdf.url = "http://cran.r-project.org/web/packages/", 
  f.zip = TRUE, f.pdf = TRUE)

Arguments

Details

This function is mainly used to download the source version of one or multiple packages. A zip version and a PDF manual can also be downloaded at the same time.

Value

Return the packge names.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

See Also

download.packages; download.file.

Examples

  download.lib(pkgs = 'erer', destdir = tempdir(), 
    f.zip = FALSE, f.pdf = TRUE)

Estimating Abnormal Return from Event Analysis

Description

Conduct an event analysis and estimate abnormal returns over time and across firms.

Usage

evReturn(y, firm, event.date, y.date = "date",
  index = "sp500", event.win = 3, est.win = 250, digits = 4, ...)

Arguments

Details

This is the core function for event analysis. It estimates a market model by firm and then calculate abnormal returns by firm and over time. The time series of stock returns have irregular time frequency because of varying trading days. Thus, the time dimension is explicitly specified as a y.date column in the data of y.

Value

Return a list object of class "evReturn" with the following components:

Methods

Two methods are defined as follows:

print:

print three selected outputs.

plot:

plot average cumulative abnormal returns from event analysis versus days in event window.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Mei, B., and C. Sun. 2008. Event analysis of the impact of mergers and acquisitions on the financial performance of the U.S. forest products industry. Forest Policy and Economics 10(5):286-294.

Sun, C., and X. Liao. 2011. Effects of litigation under the Endangered Species Act on forest firm values. Journal of Forest Economics 17(4):388-398.

See Also

evRisk

Examples

data(daEsa)

# event analysis for one firm and one event window
hh <- evReturn(y = daEsa, firm = "wpp", 
   y.date = "date", index = "sp500", est.win = 250, digits = 3,
   event.date = 19990505, event.win = 5)
hh; plot(hh)

# event analysis for many firms and one event window
hh2 <- update(hh, firm = c("tin", "wy", "pcl", "pch")); hh2

# event analysis for many firms and many event windows: need a for loop

Risk Evaluation for Event Analysis

Description

Conduct a risk analysis by firm and evaluate the change of risk before and after an event. The model used is the Captial Asset Pricing Model.

Usage

evRisk(x, m = 50, r.free = "tbill", ...)

Arguments

Details

This fits CAPM for each firm and reports the statistics for alpha, beta, and gamma. The statistics of gamma reveal the change of risk before and after the event.

Value

Return a list object of class "evReturn" with the following components:

Methods

One method is defined as follows:

print:

print selected outputs.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Mei, B., and C. Sun. 2008. Event analysis of the impact of mergers and acquisitions on the financial performance of the U.S. forest products industry. Forest Policy and Economics 10(5):286-294.

See Also

evReturn

Examples

data(daEsa)

hh <- evReturn(y = daEsa, firm = "wpp", 
   y.date = "date", index = "sp500", est.win = 250, digits = 3,
   event.date = 19990505, event.win = 5)
hh2 <- update(hh, firm = c("tin", "wy", "pcl", "pch"))

kk <- evRisk(x = hh2, m = 100, r.free="tb3m")
kk

Return the first or last part of time series data

Description

Return the first of last parts of an object of time series data.

Usage

  headts(x, n = 5, ...)
  tailts(x, n = 5, ...)

Arguments

Details

The data can be an univariate or multivariate time series data.

Value

An object like x but generally smaller.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

Examples

h1 <- ts(data=cbind(1:24), start=c(2001, 1), frequency=12)
h2 <- ts(data=cbind(1:24, 25:48), start=c(2001, 1), frequency=12)
h3 <- ts(data=cbind(1:4, 5:8, 9:12), start=c(2001, 1), frequency=4)
colnames(h2) <- c("aa", "bb")
colnames(h3) <- c("cc", "dd", "ee")
h1; h2; h3

h1; headts(h1); tailts(h1, 28)
h2; headts(h2); tailts(h2, 50)
h3; headts(h3, 2); tailts(h3); tailts(h3, 8)

data(daBed); headts(daBed); tailts(daBed)

Generate a list object with names

Description

Generate a list object with names.

Usage

listn(...)

Arguments

Details

This generates a list object by addressing the naming problem. For list, if no names are given, the list generated will have no names. In some situations, the number of individual objects is large and the names of these individual objects can be used as the names. This function addresses this need. If names are given, they will be used. If not, the names of individual objects will be used.

Value

Return a list object with names.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

Examples

y1 <- 1:10
y2 <- c("a", "b")
listn(y1, y2)
listn(y1 = y1, y2)
listn(y1 = y1, y2.rev = y2, y2, 5:8, c("d", "f"))

identical(listn(y1, y2), listn(y1 = y1, y2))        # TRUE
identical(listn(y1, y2), list(y1 = y1, y2))        # FALSE
identical(listn(y1, y2), list(y1 = y1, y2=y2))     # TRUE

List Objects with Their Sizes

Description

Show the name and memory size of objects in an R session.

Usage

lss(n = 5, pos = 1, decreasing = TRUE, order.by=c("Size", "Type")) 

Arguments

Details

This function shows the names and sizes of objects in an R session. This is useful for managing avaiable memory in an R Session.

Value

Return a dataframe object with the following columns: name, type, size, prettysize, rows, and columns.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

See Also

ls.

Examples

  lss()
  ls()

Marginal Effect for Binary Probit and Logit Model

Description

This function calculates marginal effects for a binary probit or logit model and their standard errors.

Usage

maBina(w, x.mean = TRUE, rev.dum = TRUE, digits = 3, 
  subset.name = NULL, subset.value)

Arguments

Details

Marginal effects from a binary probit or logit model is calculated. The two choices are the method of averaging effects and revising estimates for dummy variables. Marginal effects can be calculated at the mean of the independent variables (i.e., x.mean = TRUE), or as the average of individual marginal effects at each observation (i.e., x.mean = FALSE). rev.dum = TRUE allows marginal effects for dummy variables are calculated differently, instead of treating them as continuous variables.

In addition, the data set used for calculating the marginal effect can be the whole data set, or a subsect of the whole data set. In subsetting the data, a variable name and its value should be supplied. This is generally applied on a dummy variable, as shown in the example.

Value

Return a list object of class "maBina" with the following components:

Methods

One method is defined as follows:

print:

print the key output of marginal effects.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Greene, W.H. 2003. Econometric Analysis (5th ed.). Prentice Hall, New York. 1026 P.

See Also

maTrend; plot.maTrend.

Examples

data(daPe)
ma <- glm(grade ~ gpa + tuce + psi, x = TRUE, 
    data = daPe, family = binomial(link = "probit"))
    
ea <- maBina(w = ma, x.mean = TRUE, rev.dum = TRUE)
eb <- maBina(w = ma, x.mean = TRUE, rev.dum = TRUE, 
  subset.name = "psi", subset.value = 0) 
ec <- maBina(w = ma, x.mean = TRUE, rev.dum = TRUE,
  subset.name = "psi", subset.value = 1)
ea; eb; ec
eb$x; ec$x

Trend of probability values in a binary choice model

Description

This function computes the change of probability for a continuous variable based on a binary choice model (either probit or logit), and furthermore, stratifies the probability through a binary independant variable.

Usage

maTrend(q, n = 300, nam.c, nam.d, simu.c = TRUE)

Arguments

Details

For a continous variable, its probability values can be computed on the basis of a binary probit or logit model. The trend can be stratified by a binary independent variable. In addition, the standard errors of each probability series are also computed using delta method.

Value

Return a list object of class "maTrend" with the following components:

Methods

Two methods are defined as follows:

print:

print the probability output.

plot:

Plot the probability values for a continuous variable. If a strata is specified through nam.d in maTrend(), then the stratified values also are shown.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Greene, W.H. 2003. Econometric Analysis (5th ed.). Prentice Hall, New York. 1026 P.

See Also

maBina; print.maTrend; plot.maTrend.

Examples

data(daPe)
ma <- glm(grade ~ gpa + tuce + psi, x = TRUE, 
    data = daPe, family = binomial(link = "probit"))
summary(ma)

(ea <- maBina(w = ma, x.mean = TRUE, rev.dum = TRUE))
(ta <- maTrend(q = ea, nam.c = "gpa", simu.c = FALSE))
(tb <- maTrend(q = ea, nam.c = "gpa", nam.d= "psi", simu.c = TRUE, n=100))
plot(ta)
plot(tb)

Marginal Effect for Ordered Choice Model

Description

This function calculates marginal effects for an ordered chioce model and their standard errors. Either an ordered probit or logit model can be accommodated.

Usage

ocME(w, rev.dum = TRUE, digits = 3)

Arguments

Details

Marginal effects from an ordered probit or logit model is calculated. Marginal effects are calculated at the mean of the independent variables. rev.dum = TRUE allows marginal effects for dummy variables are calculated differently, instead of treating them as continuous variables. The standard errors are computed by delta method. The software of LIMDEP is used to compare and benchmark the results.

Value

Return a list object of class "ocME" with the following components:

Methods

One method is defined as follows:

print:

print the key output of marginal effects.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Greene, W.H. 2003. Econometric Analysis (5th ed.). Prentice Hall, New York. 1026 P.

See Also

ocProb

Examples

# Loading data from the MASS library
library(MASS)
data(housing) 
str(housing); head(housing)

# Fit an ordered choice model with polr from the MASS library
# This is the original specification used in MASS.
fm <- Sat ~ Infl + Type + Cont
ra <- polr(fm, data = housing, weights = Freq, Hess = TRUE, method = "probit")
rb <- polr(fm, data = housing, weights = Freq, Hess = TRUE, method = "logistic")
summary(ra); summary(rb)

# Compute the marginal effect
mea <- ocME(w = ra); mea
meb <- ocME(w = rb); meb
meb$out

Probability trend for a continuous variable in an ordered choice model

Description

This function computes the probability values for a continuous variable, based on an ordered choice model. Either an ordered probit or logit model can be used.

Usage

ocProb(w, nam.c, n = 100, digits = 3)

Arguments

Details

This function computes the probability values associated with a continous variable in an ordered probit or logit model. The standard errors of each probability value by factor level is computed using delta method. The software of LIMDEP is used to compare and benchmark the results.

Value

Return a list object of class "ocProb" with the following components:

w, nam.c, method=w$method, mean.x=colMeans(x), out, lev)

Methods

Two methods are defined as follows:

print:

print the probability output.

plot:

Plot the probability values for a continuous variable.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Greene, W.H. 2003. Econometric Analysis (5th ed.). Prentice Hall, New York. 1026 P.

See Also

ocME

Examples

# Loading data from the MASS library
library(MASS)
data(housing) 
str(housing); head(housing)

# Fit an ordered choice model with polr from the MASS library
# Note this is a fake specification
# The variable of "Freq" is included to have a 
#   continuous variable for demonstration.
fm2 <- Sat ~ Infl + Type + Cont + Freq
rc <- polr(fm2, data = housing, Hess = TRUE, method = "probit")
rd <- polr(fm2, data = housing, Hess = TRUE, method = "logistic")
summary(rc); summary(rd)

# Compute predicated probabilities with one continous variable
(fa <- ocProb(w = rc, nam.c = 'Freq', n = 300))
(fb <- ocProb(w = rd, nam.c = 'Freq', n = 300))
plot(fa)
plot(fb) 

Summary of Results from Static or Dynamic Models

Description

This summarizes the main results from AIDS models.

Usage

## S3 method for class 'aiFit'
summary(object, digits=3, ...)

Arguments

Details

This wraps up the coefficents and statistics from aiFit by equation.

Value

A data frame object with coefficients and related statistics by equation.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

See Also

aiStaFit and aiDynFit.

Examples

# see the examples for 'aiDynFit'.

Linear Equation System Estimation with Correction on Autocorrelation

Description

Fits a set of linear structural equations using Ordinary Least Squares (OLS), Weighted Least Squares (WLS), Seemingly Unrelated Regression (SUR), with the option of autocorrelation correlation.

Usage

systemfitAR(formula, method = "OLS", inst = NULL, data = list(),
  restrict.matrix = NULL, restrict.rhs = NULL, restrict.regMat = NULL, 
  pooled = FALSE, control = systemfit.control( ... ),
  AR1 = FALSE, rho.sel = c("all", "mean"), model = c("static", "dynamic"), ...)

Arguments

Details

This is a wrapper of systemfit with an addition of autocorrelation correction. It is mainly used for SUR model with autocorrelation. The main reference sources are Greene (2003), LIMDEP 9.0 manual, Judge et al. (1985), and Berndt and Savin (1975).

Value

systemfit returns a list of the class systemfit. This list contains one special object: "eq". It is a list and contains one object for each estimated equation. These objects are of the class systemfit.equation and contain the results that belong only to the regarding equation. In addition, there are four new items in the output:

The objects of the class systemfit and systemfit.equation have the following components (the elements of the latter are marked with an asterisk (*)):

Author(s)

Changyou Sun (edwinsun258@gmail.com)

References

Greene, W. H. (2003) Econometric Analysis, Fifth Edition, Prentice Hall.

Judge, George G.; W. E. Griffiths; R. Carter Hill; Helmut Luetkepohl and Tsoung-Chao Lee (1985) The Theory and Practice of Econometrics, Second Edition, Wiley.

LIMDEP 9.0 software manual.

Berndt, E.R., and N.E. Savin. 1975. Estimation and hypothesis testing in singular equation systems with autoregressive disturbances. Econometrica 43(5/6):937-957.

See Also

lm; aiStaFit; systemfit

Examples

# Check Berndt and Savin (1975) dataset

# Check Kemenda data          

Augmented-Dickey-Fuller Unit Root Test revised

Description

Augmented-Dickey-Fuller Unit Root Test revised

Usage

ur.df2(y, type = c("none", "drift", "trend"), lags = 1, 
  selectlags = c("Fixed", "AIC", "BIC"), digit = 2)

Arguments

Details

This is a modification of ur.df in the library of urca. The function was written in S4, and it is changed into S3. The lag selected by AIC or BIC is reported explicitly through lag.used in the output list. In addition, the values of AIC and BIC statistics are reported.

Value

Return an object of class ur.df2: the new outputs are lag.used, aic, and bic.

Methods

Two methods are defined as follows:

print:

print test statistics and critical values.

plot:

plot outputs.

Author(s)

Changyou Sun (edwinsun258@gmail.com)

See Also

ur.df in urca library.

Examples

# see the code for the study: Japan and China wood product imports

Output and Write a List Object

Description

Print or write its required argument z to a file

Usage

write.list(z, file, t.name = NULL, row.names = FALSE, ...) 

Arguments

Details

This function is a wrap-up of write.table. It is convenient to write a set of tables to C drive.

Value

No return value

See Also

write.table.

Examples

h1 <- ts(data=cbind(1:24), start=c(2001, 1), frequency=12)
h2 <- ts(data=cbind(1:24, 25:48), start=c(2001, 1), frequency=12)
h3 <- ts(data=cbind(1:4, 5:8, 9:12), start=c(2001, 1), frequency=4)
colnames(h2) <- c("aa", "bb")
colnames(h3) <- c("cc", "dd", "ee")
h1; h2; h3

test <- list(t1 = h1, t2 = h2, t3 = h3)

# The above list file can be saved at a specific working directory.