RIDischarge.Rmd

---
title: "RIDischarge"
author: "Miguel Leon"
date: "Monday, April 13, 2015"
output: html_document
---


```{r, echo=FALSE, message=FALSE}


# Load the project
project_directory <- "C:/Users/leonmi/Google Drive/Discharge" # library(devtools)
#project_directory <- "C:/Users/aa20/Dropbox/Manuscripts/Composite Method/composite_client"
library(ProjectTemplate)
#source(paste0(project_directory,"/","lib/loadProject.R"))
setwd(project_directory)
source("C:/Users/leonmi/Google Drive/loadest/composite_client/src/scripts/reportHelpers.R")

# Redefine project_directory now that loadProjectFromCache has cleared the old variable
project_directory <- "C:/Users/leonmi/Google Drive/Discharge" 
#project_directory <- "C:/Users/aa20/Dropbox/Manuscripts/Composite Method/composite_client"

# Set options for producing the html filee
library(knitr)
opts_chunk$set(echo=FALSE, message=FALSE, warn=FALSE)
options(warn=-1)

# Load libraries we'll use, suppressing messages via message=FALSE option to knitr chunk
library(MASS)
library(rloadest)
library(lubridate)
library(dplyr)
library(ggplot2)
library(data.table)
#RIDischargeFrom1999
insertRow <- function(existingDF, newrow, r) {
  existingDF[seq(r+1,nrow(existingDF)+1),] <- existingDF[seq(r,nrow(existingDF)),]
  existingDF[r,] <- newrow
  existingDF
}


newnames <- as.vector(t(read.csv(
  "RIDischarge.csv",
  sep=",",header=FALSE,skip=0,nrows=1,stringsAsFactors=FALSE)[1,]))

RIFullFlow <- read.csv(
  "RIDischarge.csv",
  sep=",",header=FALSE,col.names=newnames,skip=1,stringsAsFactors=FALSE)

newnames <- as.vector(t(read.csv(
  "RIDischargeFrom2006.csv",
  sep=",",header=FALSE,skip=0,nrows=1,stringsAsFactors=FALSE)[1,]))

RiFrom2006 <- read.csv(
  "RIDischargeFrom2006.csv",
  sep=",",header=FALSE,col.names=newnames,skip=1,stringsAsFactors=FALSE)
#DailyDischargeUSGS.csv

newnames <- as.vector(t(read.csv(
  "DailyDischargeUSGS.csv",
  sep=",",header=FALSE,skip=0,nrows=1,stringsAsFactors=FALSE)[1,]))

USGSDischarge <- read.csv(
  "DailyDischargeUSGS.csv",
  sep=",",header=FALSE,col.names=newnames,skip=1,stringsAsFactors=FALSE)

USGSDischarge$Date <- as.POSIXct(strptime(USGSDischarge$Date  ,format="%m/%d/%Y",tz="America/Puerto_Rico"))


RIFullFlow$DateTime <- as.POSIXct(strptime(RIFullFlow$DateTime ,format="%m/%d/%Y %H:%M",tz="America/Puerto_Rico"))


RiFrom2006$DateTime <- as.POSIXct(strptime(RiFrom2006$DateTime ,format="%m/%d/%Y %H:%M",tz="America/Puerto_Rico"))

RIFullFlow<- merge(RIFullFlow,RiFrom2006,all=TRUE)
#remove NA values 
RIFullFlow <- RIFullFlow[!is.na(RIFullFlow$USGS_cfps),]
RIFullFlow<- RIFullFlow[order(RIFullFlow$DateTime),]
  

RIFullFlow$num <- 1
min15pyr <- with(RIFullFlow, tapply(num, YEAR, sum))

```

15 minute (or 10 min) records per year

```{r, echo=FALSE, message=FALSE}

min15pyr

dt <- as.POSIXlt("1986-10-01 00:15:00")
dt <- dt -900
newrow <- list(1986,1,as.POSIXct(strptime(dt, format="%Y-%m-%d",tz="America/Puerto_Rico")),NA,NA,1986,1) #%H:%M:%S

#,origin = "1900-01-01"
#"1999-10-01 00:00:00","%Y-%m-%d %H:%M:%S",tz="America/Puerto_Rico"
#USGS_Sonadora_IDA2 <- rbind(USGS_Sonadora_IDA, newrow)
RIFullFlow<- insertRow(RIFullFlow, newrow, 1)

RIFullFlow$diffsec <- c(NA, tail(RIFullFlow$DateTime, -1) - head(RIFullFlow$DateTime, -1))
RIFullFlow$diffMin <- RIFullFlow$diffsec / 60

RIFullFlow$diffsec[RIFullFlow$diffsec > 900] <- 900
RIFullFlow$diffMin[RIFullFlow$diffMin > 15] <- 15

#some of these are 5 minute not 15 minute need to make sure the cfps for each record represents the right amount of time. 
totalCF15mCurr<- with(RIFullFlow, tapply(((USGS_cfps*diffsec)), YEAR, sum))
totalCF15m2Curr <- totalCF15mCurr /with(RIFullFlow, tapply(((diffsec)), YEAR, sum))

```

mean discharge per year 

```{r, echo=FALSE, message=FALSE}

totalCF15m2Curr
write.csv(totalCF15m2Curr,
    "./RIMeanAnnualDischarge.csv",na="NA",,row.names=TRUE)

```

plot log discharge 2010 

```{r, echo=FALSE, message=FALSE}



#agggregate by day 

```

calculate mean daily discharge

```{r, echo=FALSE, message=FALSE}

RIFullFlow$Day<- as.POSIXct(strptime(RIFullFlow$DateTime,format="%Y-%m-%d",tz="America/Puerto_Rico"))

#FILL IN DISCHARGE MEAN FOR ALL TIMES 
RIFullFlow<- RIFullFlow %>%
  group_by(Day) %>% 
  mutate(cfsps_day_mean=mean(USGS_cfps,na.rm=TRUE))
days<- list(RIFullFlow$Day)
RIFullFlowDaily<- aggregate(RIFullFlow,by=days, FUN=head, 1)

#write daily discharge to a file 
write.csv(RIFullFlowDaily,
    "./RIDailyExport.csv",na="NA",,row.names=FALSE)

#Q12014day<- transform(RIFullFlow, YR=as.numeric(format(Date, "%Y")))
years<- list(RIFullFlowDaily$YEAR)
RIFullFlowMeanDailytoYr<- aggregate(RIFullFlowDaily$cfsps_day_mean,by=years, FUN=sum)



startDate <- as.POSIXct("2013-01-01") 
endDate <- as.POSIXct("2014-01-01")

RIFullFlowDaily2<- RIFullFlowDaily[RIFullFlowDaily$DateTime >  startDate 
                        & RIFullFlowDaily$DateTime < endDate,]

USGSDischarge2<- USGSDischarge[USGSDischarge$Date >  startDate 
                        & USGSDischarge$Date < endDate,]

  ggplot() + ggtitle("RI log Discharge CFS 2013") + 
    geom_line(data= RIFullFlowDaily2, aes(x=DateTime, y=log(USGS_cfps) , color="Mean Daily Discharge"))+
   geom_line(data= USGSDischarge2, aes(x=Date, y=log(cfs.1) , color="Mean Daily Discharge USGS"))

```


Daily mean CFS Stats
1st is SD

```{r, echo=FALSE, message=FALSE}

sd(RIFullFlowDaily$cfsps_day_mean)
summary(RIFullFlowDaily$cfsps_day_mean)

```


Daily mean CFS Stats from USGS data 
1st is SD

```{r, echo=FALSE, message=FALSE}

sd(USGSDischarge$cfs.1)
summary(USGSDischarge$cfs.1)

```


Daily mean CFS Stats 2014
1st is SD

```{r, echo=FALSE, message=FALSE}

sd(RIFullFlowDaily2$cfsps_day_mean)
summary(RIFullFlowDaily2$cfsps_day_mean)

```


Daily mean CFS Stats from USGS data 2014
1st is SD

```{r, echo=FALSE, message=FALSE}

sd(USGSDischarge2$cfs.1)
summary(USGSDischarge2$cfs.1)

```

days per year

```{r, echo=FALSE, message=FALSE}


daypYr <- with(RIFullFlowDaily, tapply(num, YEAR, sum))
daypYr
```

calculate annual mean discharge from daily average discharge

```{r, echo=FALSE, message=FALSE}
write.csv(daypYr,
    "./RIDaysPYear.csv",na="NA",row.names=TRUE)

RIFullFlowMeanDailytoYr$x <- RIFullFlowMeanDailytoYr$x /daypYr
RIFullFlowMeanDailytoYr

write.csv(RIFullFlowMeanDailytoYr,
    "./meanYearlyDischargeFromMeanDaily.csv",na="NA",row.names=TRUE)




```

find days per month

```{r, echo=FALSE, message=FALSE}


RIFullFlowMonthly<- transform(RIFullFlowDaily, Month=format(DateTime, "%Y-%m"))

daypMonth <- with(RIFullFlowMonthly, tapply(num, Month, sum))
daypMonth

write.csv(daypMonth,
    "./RIDaysPMonth.csv",na="NA",,row.names=TRUE)

RIFullFlowMonthly<- RIFullFlowMonthly %>%
  group_by(Month) %>% 
  mutate(cfsps_month_mean=mean(cfsps_day_mean,na.rm=TRUE))

months<- list(RIFullFlowMonthly$Month)
RIFullFlowMonthly<- aggregate(RIFullFlowMonthly,by=months, FUN=head, 1)

write.csv(RIFullFlowMonthly,
    "./RIMonthlyExport.csv",na="NA",,row.names=FALSE)



```

Monthly mean CFS Stats
1st is SD

```{r, echo=FALSE, message=FALSE}

sd(RIFullFlowMonthly$cfsps_month_mean)
summary(RIFullFlowMonthly$cfsps_month_mean)