## ----setup, include=FALSE-----------------------------------------------------
knitr::opts_chunk$set(echo = TRUE)


## ----get----------------------------------------------------------------------
ff <- agro::get_data_from_uri("hdl:11529/11111", ".")
ff


## ----read---------------------------------------------------------------------
ff[1]
d <- read.csv(ff[1], stringsAsFactors = FALSE)
dim(d)


## ----colnames-----------------------------------------------------------------
colnames(d)


## ----table--------------------------------------------------------------------
table(d$Year)


## ----y------------------------------------------------------------------------
yield <- d[, c("Yield_A", "Yield_B", "Yield_C")]
head(yield)


## ----y2, error=TRUE-----------------------------------------------------------
y_avg <- rowMeans(yield)


## ----y4-----------------------------------------------------------------------
sapply(yield, class)


## ----y6-----------------------------------------------------------------------
y <- apply(yield, 2, as.numeric)


## ----y16----------------------------------------------------------------------
na_rows <- apply(is.na(y), 1, any)
table(na_rows)
nas <- which(na_rows)
head(nas)
yield[nas[1:5], ]


## ----y20----------------------------------------------------------------------
table(d$Year, na_rows)


## ----y30----------------------------------------------------------------------
yield[yield == "-"] <- NA
y <- apply(yield, 2, as.numeric)


## ----y40, error=TRUE----------------------------------------------------------
y_avg <- rowMeans(y, na.rm=TRUE)


## ----y45----------------------------------------------------------------------
d$yield <- y_avg


## ----y50----------------------------------------------------------------------
plot(sort(d$yield))


## ----y51----------------------------------------------------------------------
tapply(d$yield, d$Year, mean, na.rm=TRUE)


## ----y55----------------------------------------------------------------------
boxplot(yield~Year, data=d)


## ----m1-----------------------------------------------------------------------
ff[2]
md <- read.csv(ff[2], stringsAsFactors = FALSE, header=FALSE)
head(md)


## ----m11----------------------------------------------------------------------
colnames(md) <- c("var", "value")


## ----m15----------------------------------------------------------------------
i <- grep("Yield", md$var)
i
md$var[i]


## ----m21----------------------------------------------------------------------
md <- apply(md, 2, trimws)
# apply returns a matrix, let's keep a data.frame (always with stringsAsFactors=FALSE)
md <- data.frame(md, stringsAsFactors = FALSE)


## ----m31----------------------------------------------------------------------
md$value[i]


## ----m41----------------------------------------------------------------------
ymd <- gsub("Grain weight at spot .\n", "", md$value[i])
ymd <- unique(ymd)
ymd


## ----m51----------------------------------------------------------------------
cat(ymd)


## ----m61----------------------------------------------------------------------
units <- data.frame(Year=2013:2016, area=c(25, 2*2.5 ,1 , 4), tokg=c(1,1,1000,1000))
units


## ----ym1----------------------------------------------------------------------
dim(d)
dm <- merge(d, units, by="Year")
dim(dm)


## ----ym2----------------------------------------------------------------------
head(dm[, c("Year", "yield", "area", "tokg")])
tail(dm[, c("Year", "yield", "area", "tokg")])
table(dm[, c("Year", "area")])


## ----ym10---------------------------------------------------------------------
dm$yha <- (10000 / dm$area) * (dm$yield / dm$tokg)


## ----ym15---------------------------------------------------------------------
tapply(dm$yha, dm$Year, mean, na.rm=TRUE)


## ----ym21---------------------------------------------------------------------
dm$tokg[dm$Year==2015] <- 1
dm$yha <- (10000 / dm$area) * (dm$yield / dm$tokg)


## ----ym31---------------------------------------------------------------------
tapply(dm$yha, dm$Year, mean, na.rm=TRUE)


## ----ym41---------------------------------------------------------------------
dm$Gr_Moisture[dm$Gr_Moisture == "-"] <- NA
dm$Gr_Moisture <- as.numeric(dm$Gr_Moisture)
table(dm$Year, is.na(dm$Gr_Moisture))
boxplot(Gr_Moisture~Year, data=dm, col=rainbow(6), ylab="Yield (kg/ha)", cex.axis=1.25, cex.lab=1.25)


## ----ym42---------------------------------------------------------------------
moisture <- aggregate(dm[, "Gr_Moisture", drop=FALSE], dm[, "Year", drop=FALSE], mean, na.rm=TRUE)
colnames(moisture)[2] <- "avg_moisture"
moisture


## ----ym44---------------------------------------------------------------------
dim(dm)
dm <- merge(dm, moisture, by="Year")
dim(dm)


## ----ym45---------------------------------------------------------------------
i <- is.na(dm$Gr_Moisture)
dm$Gr_Moisture[i] <- dm$avg_moisture[i]
table(is.na(dm$Gr_Moisture))
boxplot(Gr_Moisture~Year, data=dm, col=rainbow(6), ylab="Yield (kg/ha)", cex.axis=1.25, cex.lab=1.25)


## -----------------------------------------------------------------------------
dm$yha <- 12 * dm$yha / dm$Gr_Moisture 
boxplot(yha~Year, data=dm, col=rainbow(6), ylab="Yield (kg/ha)", cex.axis=1.25, cex.lab=1.25)


## ----f1-----------------------------------------------------------------------
fert <- c("DAP_Total", "MoP_Total", "Urea_Total", "Urea_1_Top", "MoP_1_Top", "Urea_2_Top", "MoP_2_Top", "Urea_3_Top", "MoP_3_Top")
md[md$var %in% fert, ]


## ----f2-----------------------------------------------------------------------
aggregate(dm[, fert], dm[,"Year",drop=FALSE], max, na.rm=TRUE)


## ----f10----------------------------------------------------------------------
i <- dm$Year %in% c(2013, 2014)
sum(i)
dm$Urea_Total[i] <- rowSums(dm[i, c("Urea_1_Top", "Urea_2_Top", "Urea_3_Top")])

j <- dm$Year == 2014
sum(j)
dm$MoP_Total[j] <- rowSums(dm[j, c("MoP_1_Top", "MoP_2_Top", "MoP_3_Top")])


## ----f20----------------------------------------------------------------------
df <- dm[dm$Year != 2016, ]
aggregate(df[, fert[1:3]], df[,"Year",drop=FALSE], mean, na.rm=TRUE)