Agricultural Goods Price Forecast
Introduction
Agriculture is referred to as the production, processing and distribution of agricultural goods. It can be also the main source income for most developing countries. In addition to providing food and raw material, agriculture also provides employment opportunities to a very large percentage of the population.
Objetive
We are going to make use of our analytical knowledge to analyze and forecast agricultural goods prices of Rwanda, a developing country.
Data
For this project we’ll use the United Nations Humanitarian Data Exchange Global Food Price Database that can be found here.
The database covers agricultural goods like beans, rice, maize, fish, and sugar for 76 countries and some 1,500 markets. The data is updated weekly and goes back as far as 1992 for a few countries, although many countries started reporting from 2003 or thereafter.
library(dplyr)
library(readr)
library(lubridate)
library(knitr)
library(ggplot2)
library(magrittr)
library(forecast)
potato <- read_csv("Potato.csv", col_types = cols_only(adm1_name = col_character(),
mkt_name = col_character(), cm_name = col_character(),
mp_month = col_integer(), mp_year = col_integer(),
mp_price = col_number()))
potato_rename <- rename(potato, "region" = "adm1_name", "market" = "mkt_name",
"commodity_kg" = "cm_name", "month" = "mp_month", "year" = "mp_year",
"price_rwf" = "mp_price")
glimpse(potato_rename)Rows: 4,320
Columns: 6
$ region <chr> "$West/Iburengerazuba", "$West/Iburengerazuba", "$West/Ib~
$ market <chr> "Birambo", "Birambo", "Birambo", "Birambo", "Birambo", "B~
$ commodity_kg <chr> "Potatoes (Irish)", "Potatoes (Irish)", "Potatoes (Irish)~
$ month <int> 11, 12, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 1, 2, 3, 4~
$ year <int> 2010, 2010, 2011, 2011, 2011, 2011, 2011, 2011, 2011, 201~
$ price_rwf <dbl> 157.0000, 133.3333, 96.5000, 97.0000, 107.8000, 125.5000,~Data cleaning
As usually raw data doesn’t arrives as we would want, so we have to make some arrengments and manipulate the data so we can work with it.
potato_clean <- potato_rename %>%
mutate(date = ymd(paste(year, month, '01'))) %>%
select(-year, -month)
head(potato_clean)# A tibble: 6 x 5
region market commodity_kg price_rwf date
<chr> <chr> <chr> <dbl> <date>
1 $West/Iburengerazuba Birambo Potatoes (Irish) 157 2010-11-01
2 $West/Iburengerazuba Birambo Potatoes (Irish) 133. 2010-12-01
3 $West/Iburengerazuba Birambo Potatoes (Irish) 96.5 2011-01-01
4 $West/Iburengerazuba Birambo Potatoes (Irish) 97 2011-02-01
5 $West/Iburengerazuba Birambo Potatoes (Irish) 108. 2011-03-01
6 $West/Iburengerazuba Birambo Potatoes (Irish) 126. 2011-04-01Now we have our cleaned data, an can follow the same process for other food types, we have many options:
- Beans
- Chili
- Cassava
- Oranges
- Maize
- Peas
- Sorghum
- Tomatoes
So, let’s repeat the same process over and over! Just kidding, even though you can do that it is very unconvenient so we are going to build some functions to make our lives easier. However, we are going to continue with our potato as example of how you could do this analysis if it was just about one topic (potatoes), but you will find it is very much convenient to write functions in the end.
First of all, let’s get a function that can read our data and define each column type.
read_data <- function(commodity){
data_file <- paste0(commodity, '.csv')
prices <- read_csv(
data_file,
col_types = cols_only(
adm1_name = col_character(),
mkt_name = col_character(),
cm_name = col_character(),
mp_month = col_integer(),
mp_year = col_integer(),
mp_price = col_double()))
prices_renamed <- prices %>%
rename(
region = adm1_name,
market = mkt_name,
commodity_kg = cm_name,
month = mp_month,
year = mp_year,
price_rwf = mp_price
)
prices_renamed %>%
mutate(
date = ymd(paste(year, month, "01"))
) %>%
select(-month, -year)
}
pea <- read_data("Pea")
glimpse(pea)Rows: 1,893
Columns: 5
$ region <chr> "$West/Iburengerazuba", "$West/Iburengerazuba", "$West/Ib~
$ market <chr> "Birambo", "Birambo", "Birambo", "Birambo", "Birambo", "B~
$ commodity_kg <chr> "Peas (fresh)", "Peas (fresh)", "Peas (fresh)", "Peas (fr~
$ price_rwf <dbl> 403.5000, 380.0000, 277.5000, 450.0000, 450.0000, 375.000~
$ date <date> 2011-01-01, 2011-02-01, 2011-04-01, 2011-05-01, 2011-06-~Exploratory Data Analysis
One first good step when you have historical data and want to start the analysis, its to plot the data you have gather. So, let’s see how the prices of potatoes has change over time.
ggplot(data = potato_clean, aes(x = date, y = price_rwf, group = market)) +
geom_line(alpha = 0.2) +
ggtitle("Potato Price Across the Years (2013-2016)") +
labs(x='Date', y='Price')
As you can see, there is a increasing trend on the prices of potatoes. More particularly, it seems that the prices relays on the season: for each year the price seems to drop at December-January and rise through the year until a peak around August.
We want to make similiar plots to the one above, so let’s wrap the plotting code into a function.
price_plot <- function(prices, commodity){
title <- paste(commodity, "Price Across the Years (2011-2016)")
prices %>%
ggplot(aes(date, price_rwf, group = market)) +
geom_line(alpha = 0.2) +
ggtitle(title) +
labs(x='Date', y='Price')
}
price_plot(pea, "Pea")
Forecasting
Yes, it can be really helpful to get insights about the present status of particular circumstances, but what can be way more exciting its to get insight of how the situation may change n the future, in other words, make predictions.
In this situation we are working with food prices, so, we will take the median price across markets and analyze the resulting time series.
The reason we are taking the madian and not the mean it is because, by looking at the plots and the big spike in the price, we can conclude That probably these spikes probably indicate a logistic problem. Whether the food wasn’t easily available at the market, or the harvest season wasn’t good, the consequence of these outliers is that it is a bad idea to use the mean price of each time point. Instead, the median makes more sense since it is robust against outliers.
potato_summary <- potato_clean %>%
group_by(date) %>%
summarize(median_price_rwf = median(price_rwf))
head(potato_summary)# A tibble: 6 x 2
date median_price_rwf
<date> <dbl>
1 2008-01-01 97.5
2 2008-02-01 100
3 2008-03-01 95
4 2008-04-01 96.2
5 2008-05-01 95
6 2008-06-01 110 potato_ts <- potato_summary %$%
ts(median_price_rwf, start = c(year(min(date)), month(min(date))),
end = c(year(max(date)), month(max(date))), frequency = 12)
potato_ts Jan Feb Mar Apr May Jun Jul Aug
2008 97.5000 100.0000 95.0000 96.2500 95.0000 110.0000 116.6667 125.0000
2009 120.0000 122.5000 130.0000 131.2500 135.0000 124.3125 125.8333 144.2500
2010 109.6875 113.5000 131.2500 132.0833 140.4167 147.3750 142.5000 161.5000
2011 105.7000 108.1750 118.8750 145.0143 148.6667 148.0500 137.4048 137.2619
2012 150.7500 175.2500 186.0139 186.2500 182.5000 162.7500 179.1250 196.9643
2013 154.3333 157.0000 171.2500 187.5000 177.0000 202.2500 210.0000 233.1875
2014 138.3333 158.7500 186.2500 198.2500 191.0000 189.3333 182.5000 187.6191
2015 136.2500 157.6071 178.0000 190.2778 179.3750 168.3333 180.0000 202.1250
Sep Oct Nov Dec
2008 136.2500 130.0000 127.5000 114.3750
2009 181.2500 170.0000 150.2500 112.0000
2010 182.4000 162.5000 151.5000 122.5000
2011 141.6667 144.2000 133.1750 141.5000
2012 226.5000 203.5000 169.2500 144.0000
2013 241.3333 237.5000 176.7083 140.0000
2014 200.0000 183.1310 150.0000 133.9286
2015 223.5000 217.5000 216.1250 190.0000time_series <- function(prices){
prices_summarized <- prices %>%
group_by(date) %>%
summarize(median_price_rwf = median(price_rwf))
time_series <- prices_summarized %$%
ts(
median_price_rwf,
start = c(year(min(date)), month(min(date))),
end = c(year(max(date)), month(max(date))),
frequency = 12
)
}
pea_ts <- time_series(pea)
pea_ts Jan Feb Mar Apr May Jun Jul
2011 561.6667 700.0000 958.0000 710.0000 591.5000 597.8572 666.3572
2012 655.0000 950.0000 1272.1667 1166.0000 945.8750 822.3333 714.2857
2013 668.7500 781.6334 829.9875 975.0000 908.2500 789.9444 806.8000
2014 695.5000 1025.0000 1166.6250 1083.2500 825.0000 816.6667 809.5714
2015 800.0000 1066.6667 1100.0000 1051.8889 950.0000 873.6667 804.1250
Aug Sep Oct Nov Dec
2011 758.5000 938.8333 1506.2500 787.5000 548.9375
2012 788.1250 990.7222 1413.7500 964.2619 661.8571
2013 1000.0000 1162.4583 1316.7500 916.6667 623.8571
2014 1000.0000 1000.0000 1666.6667 700.0000 633.3333
2015 900.0000 1166.6667 1550.0000 1066.6667 802.1250potato_forecast <- forecast(potato_ts)
potato_forecast## Point Forecast Lo 80 Hi 80 Lo 95 Hi 95
## Jan 2016 190.0093 171.35706 208.6615 161.48317 218.5354
## Feb 2016 202.6099 174.14582 231.0740 159.07783 246.1420
## Mar 2016 220.0317 181.72222 258.3413 161.44238 278.6211
## Apr 2016 231.5932 184.48380 278.7026 159.54559 303.6408
## May 2016 226.2626 174.20438 278.3209 146.64641 305.8789
## Jun 2016 229.1587 170.73454 287.5829 139.80665 318.5108
## Jul 2016 230.8787 166.57270 295.1848 132.53113 329.2263
## Aug 2016 251.1739 175.53815 326.8096 135.49902 366.8487
## Sep 2016 279.3573 189.13187 369.5827 141.36943 417.3451
## Oct 2016 262.7887 172.33073 353.2467 124.44516 401.1323
## Nov 2016 236.0485 149.89274 322.2042 104.28465 367.8123
## Dec 2016 205.0924 126.05584 284.1290 84.21640 325.9684
## Jan 2017 205.0036 121.88813 288.1190 77.88948 332.1177
## Feb 2017 218.4941 125.58323 311.4050 76.39917 360.5891
## Mar 2017 237.1698 131.67270 342.6669 75.82591 398.5137
## Apr 2017 249.5154 133.68437 365.3465 72.36711 426.6638
## May 2017 243.6602 125.85363 361.4667 63.49061 423.8297
## Jun 2017 246.6667 122.68387 370.6496 57.05130 436.2822
## Jul 2017 248.4066 118.81644 377.9967 50.21556 446.5976
## Aug 2017 270.1226 124.07681 416.1684 46.76484 493.4804
## Sep 2017 300.3005 132.25584 468.3452 43.29837 557.3027
## Oct 2017 282.3675 119.02591 445.7092 32.55807 532.1770
## Nov 2017 253.5265 102.08787 404.9651 21.92111 485.1319
## Dec 2017 220.1852 84.51341 355.8570 12.69310 427.6773autoplot(potato_forecast, main = 'Potato Price Forecast', xlab='Date', ylab='Median Price')
After some data manipulation we have the forecast at last. But do we know if we can trust in this forecast? Well, recall that we inquire that the data depends of the season (low prices at December-January, and a high peak at August). Consequently, a good forecast should show a similar shape throughout the seasons.
plot_forecast <- function(time_series, commodity){
price_forecast <- forecast(time_series)
autoplot(price_forecast, main = paste(commodity, 'Price Forecast'), xlab='Date', ylab='Median Price')
}
plot_forecast(pea_ts, "Pea")
Conclusion and Recommendations
To run a forecasting model we need to convert data into a time series. This can be an usual tool to gather insights of not just the current status of a situation, but to get a view of the possible future it may have.
As shown, there was a lot of effort writing all that code to just analyze the potato data. Fortunately and by good practices, wrapping all the analysis into functions, we could easily reproduce the job with another type of food. Remember there is still more that that can be analyzed, just take the functions we already wrote and do a report.
Example: Sorghum Forecasting
commodity <- "Sorghum"
sorghum <- read_data(commodity)
price_plot(sorghum, commodity)
sorghum_ts <- time_series(sorghum)
plot_forecast(sorghum_ts, commodity)