Mini Project 4 log

[zrubinstein]

maple_sap <- maple_sap %>% select(when, sap) %>% filter(!is.na(sap))

[zrubinstein]

mac <- etl("macleish") %>% etl_update()

whately <- mac %>% tbl("whately") %>% collect(n = Inf) %>% select(when, temperature)

orchard <- mac %>% tbl("orchard") %>% collect(n = Inf) %>% select(when, temperature)

[jaatin]

whately <- whately %>% separate(col = when, into = c("when", "time"), sep = "T", remove = TRUE)

[zrubinstein]

weather1 <-inner_join(whately, orchard, by="when")

[zrubinstein]

weather1<-weather1 %>% separate(col = when, into = c("when", "time"), sep = "T", remove = TRUE) weather1<-weather1 %>% rename(whately=temperature.x, orchard = temperature.y) %>%

weather1<-weather1 %>%
filter(time=="00:00:00Z"| time=="04:00:00Z" | time== "08:00:00Z" | time=="12:00:00Z" | time== "16:00:00Z" | time== "20:00:00Z")

[zrubinstein]

weather1<-mutate(weather1, temp=(orchard+whately)/2)

[jaatin]

whately <- mac %>%
  tbl("whately") %>%
  collect(n = Inf) %>%
  select(when, temperature)

orchard <- mac %>%
  tbl("orchard") %>%
  collect(n = Inf) %>%
  select(when, temperature)

weather <- inner_join(whately, orchard, by="when") %>%
  separate(col = when, into = c("when", "time"), sep = "T", remove = TRUE) %>%
  rename(whately=temperature.x, orchard = temperature.y)

weather <- weather %>%
  mutate(temp = (orchard+whately)/2) %>%
  select(when, time, temp)

weather_day <- weather %>%
  filter(time== "08:00:00Z" | time=="12:00:00Z" | time== "16:00:00Z" | time== "20:00:00Z") %>%
  group_by(when) %>%
  summarize(avg_day_temp = mean(temp))

weather_night <- weather %>%
  filter(time=="00:00:00Z"| time=="04:00:00Z" | time== "16:00:00Z" | time== "20:00:00Z") %>%
  group_by(when) %>%
  summarize(avg_night_temp = mean(temp))

weather_all <- inner_join(weather_day, weather_night, by="when")

[zrubinstein]

'##Denote Day Temp and Night Temp weather_day <- weather %>% filter(time== "08:00:00Z" | time=="12:00:00Z" | time== "16:00:00Z" | time== "20:00:00Z") %>% group_by(when) %>% summarize(avg_day_temp = mean(temp))

weather_night <- weather %>% filter(time=="00:00:00Z"| time=="04:00:00Z" | time== "16:00:00Z" | time== "20:00:00Z") %>% group_by(when) %>% summarize(avg_night_temp = mean(temp))

Join and clean up

weather_all <- inner_join(weather_day, weather_night, by="when") weather_all<-mutate(weather_all, when=as.Date(when))

Reorganize data frame and rejoin

melted_day<-select(weather_all, avg_day_temp, when) melted_day<-melt(melted_day, id.vars=c("when"), variable.name="ToD", value.name="temperature")

melted_night<-select(weather_all, avg_night_temp, when) melted_night<-melt(melted_night, id.vars=c("when"), variable.name="ToD", value.name="temperature")

full<-rbind(melted_day, melted_night) full<-arrange(full, when)'

[zrubinstein]

need to include library(reshape2) to melt

[jkhamilton]

Spacial Data of Maple Forest -

bechtel <- data.frame(lat = 42.449167, lon = -72.679389)

forest_pal <- colorFactor("Greens", macleish_layers[["forests"]]$Sheet1__Na == 9)

leaflet() %>%

Base groups

addTiles(group = "OpenStreetMap") %>% addProviderTiles("Esri.WorldTopoMap", group = "Topography") %>% addProviderTiles("Esri.WorldImagery", group = "Satellite") %>% addProviderTiles("Stamen.TonerLite", group = "Toner Lite") %>%

Boundaries of Property

addPolygons(data = macleish_layers[["boundary"]], weight = 1, fillOpacity = 0.01, group = "Boundaries") %>%

Natural elements, filtering for Sugar Maple Forest

addPolygons(data = macleish_layers[["forests"]], color = ~forest_pal(Sheet1Na == 9), weight = 0.1, fillOpacity = 0.2, popup = ~ Sheet1Na == 9, group = "Natural") %>%

Layers control

addLayersControl( baseGroups = c("OpenStreetMap", "Topography", "Satellite", "Toner Lite"), overlayGroups = c("Boundaries", "Natural"), options = layersControlOptions(collapsed = FALSE) ) %>% setView(lat = bechtel$lat, lng = bechtel$lon, zoom = 15) m

Including data frame that we pulled data from ("forests" within MacLeish spatial data)

summary(macleish_layers[["forests"]])

[jaatin]

temp_plot <- ggplot(data = full) +
  geom_point(aes(x = when, y = temperature, color = ToD), alpha = 0.5) +
  geom_line(aes(x = when, y = 0))

temp_plot

[zrubinstein]

Add Sap Data

full2<- inner_join(full, maple_sap)

Filter by Year

full3<-full2 %>% separate(col = when, into = c("year", "month", "day"), sep = "-", remove = TRUE) yr2015<-filter(full3, year=="2015")

[jkhamilton]

Problem Statement: We are examining potential start and end dates for sugaring season. Temperatures must be above freezing during the day and below freezing at night, in order for sap to run. By combining weather data with maple_sap data, the follow graphics illustrate our findings about the sugaring season.

[zrubinstein]

Filter by Year

full3<-full2 %>% separate(col = when, into = c("year", "month", "day"), sep = "-", remove = TRUE) full4<- full3 %>%

yr2015<-filter(full3, year=="2015") yr2016<-filter(full3, year=="2016") yr2017<-filter(full3, year=="2017")

##Visualization
```{r}
temp_plot <- ggplot(data = full) +
  geom_point(aes(x = when, y = temperature, color = ToD), alpha = 0.5) +
  geom_line(aes(x = when, y = 0))
temp_plot

ggplot(data=yr2015) + geom_point(aes(x=day, y=sap))

ggplot(data=yr2016) + geom_point(aes(x=day, y=sap))

ggplot(data=yr2017) + geom_point(aes(x=day, y=sap))

[jkhamilton]

In New England, farmers are generally out in the woods from late February through early April tapping maple trees for their sap. It takes up to 40 gallons of sap from a tree to make just one gallon of syrup (Source 1). In 2016, Massachusetts had 315,000 taps with a yield per tap of 0.244 gallons and produced 77,000 gallons of syrup.

Source 1: http://www.discovernewengland.org/things-do/new-englands-maple-sugaring-season Source 2: https://www.nass.usda.gov/Statistics_by_State/New_England_includes/Publications/Current_News_Release/2016/Maple.pdf

[jkhamilton]

Actionable Recommendation: Our suggestion is that Smith taps the maple trees between and in order to obtain the maximum yields.

[jaatin]

Visualizations:

temp_plot <- ggplot(data = full) +
  geom_point(aes(x = when, y = temperature, color = ToD), alpha = 0.5) +
  geom_line(aes(x = when, y = 0))

temp_plot

## October to April
oct15_apr16 <-full %>%
  filter(grepl("(2015-(10|11|12))|(2016-(01|02|03|04))",when))

oct15_apr16_plot <-ggplot(data = oct15_apr16) +
  geom_point(aes(x = when, y = temperature, color = ToD), alpha = 0.5) +
  geom_line(aes(x = when, y = 0))

oct15_apr16_plot

[zrubinstein]

february 1 - april 30

max and min more important than average -- something in the future; maybe better if average is low (when did i get below freezing and how long did it stay there)

orchard is 30ft high as oppose to whately (80ft; above canopy)

soil moisture probes; maybe soil temperature probes

[jaatin]

## Februrary to April
feb_apr <-full %>%
  filter(grepl("(2016-(02|03|04))",when))

feb_apr_plot <-ggplot(data = feb_apr) +
  geom_point(aes(x = when, y = temperature, color = ToD), alpha = 0.5) +
  geom_line(aes(x = when, y = 0))

feb_apr_plot

[jaatin]

## Combine month and day columns in full3 df
full3_date <- full3 %>%
  mutate(month_day = paste(month, day, sep = '-')) %>%
  select(year, month_day, ToD, temperature, sap)

[jkhamilton]

(to be added to the map)

in the future: it would be extremely beneficial to have, within the "forest" Macleish package, marked trees that are being tapped. Then within our map we could annotate where each tree is being tapped and look more into temperatures at those specific points and project future tapping locations.

[jkhamilton]

bechtel <- data.frame(lat = 42.449167, lon = -72.679389)

forest_pal <- colorFactor("Greens", macleish_layers[["forests"]]$Sheet1__Na == 9)

leaflet() %>%
  ## Base groups
  addTiles(group = "OpenStreetMap") %>%
  addProviderTiles("Esri.WorldTopoMap", group = "Topography") %>%
  addProviderTiles("Esri.WorldImagery", group = "Satellite") %>%
  addProviderTiles("Stamen.TonerLite", group = "Toner Lite") %>%

  ## Boundaries of Property
  addPolygons(data = macleish_layers[["boundary"]], weight = 1, fillOpacity = 0.01, group = "Boundaries") %>%

  ## Natural elements, filtering for Sugar Maple Forest
  addPolygons(data = macleish_layers[["forests"]], 
              color = ~forest_pal(Sheet1__Na == 9), weight = 0.1, 
              fillOpacity = 0.2,
              popup = ~ Sheet1__Na == 9, group = "Natural") %>%

  ## Layers control
  addLayersControl(
    baseGroups = c("OpenStreetMap", "Topography", "Satellite", "Toner Lite"),
    overlayGroups = c("Boundaries", "Natural"),
    options = layersControlOptions(collapsed = FALSE)
  ) %>%
  setView(lat = bechtel$lat, lng = bechtel$lon, zoom = 15)
m

##Including data frame that we pulled data from ("forests" within MacLeish spatial data)
summary(macleish_layers[["forests"]])

##in the futurue: it would be extremely beneficial to have, within the "forest" Macleish package, marked trees that are being tapped. Then within our map we could annotate where each tree is being tapped and look more into temperatures at those specific points and project future tapping locations.