Data Visualization - Weather Project

Dataset

The datasets, results-milan.csv and results-global.csv were originally provided by Udacity. The datasets include 266 global temperatures from 1750 until 2015 as well as 271 temperatures of Milan collected between 1743 and 2013. The main features are of course the year and the average temperature for a whole year. To plot the moving averages I created the haskell programm weatherComparator which reads in both csv files. The moving averages of both regions are calculated based on an interval of 10 years. If a year had no temperature then this row was dropped. The calculated moving averages were written to the csv file moving-average-weather-data.csv. Last but not least the haskell program plotted the csv file which showed the moving average global temperatures vs. the moving average temperatures of Milan.

Summary of Findings

In general, the moving average temperatures of Milan are lower compared to the global moving averages.
Both categories, i.e. 'global' and 'Milan', show the same trend that the average temperature has been rising, especially since around 1945.

Review

WeatherComparator

The WeatherComparator is a program written in Haskell. It creates the moving averages based on two csv files that need to have the headings year and avg_temp (see examples in the data folder. Please note that the source csv files may contain additional headings but these two are a must. What's more, this program will only work if the binary weatherComparator together with the data folder are in the same directory. The csv files of both regions as well as the csv file containing the moving averages need to be in the data folder. That's why you'll only need to enter the names of the csv files but not their paths.

Usage

$ weatherComparator [-h] [-c csv-file1] [-c csv-file2] [-d link] [-t] [-r region1] [-r region2]

Option	Argument	Long Option	Option Description
-v		--verbose	chatty output on stderr
-V, -?		--version	show version number
-q		--quit	end program
-c	FILE	--csv=FILE	csv file for temperature comparison
-d	LINK	--chart=LINK	create a diagram
-t		--test	create a diagram based on test set
-r	REGION	--region=REGION	the name of a region
-h		--help	Show this help message

First of all, you can run this program by generating the csv file 'moving-average-weather-data.csv' as well as the htmlfile 'moving-averages.html' containing the diagram by using the testset:

$ weatherComparator -t -v

Another option is to generate the csv file 'moving-average-weather-data.csv' based on your data, for example:

$ weatherComparator -v -r "San Francisco" -r Milan -c results-milan.csv -c results-sanfrancisco.csv

The third option is to generate the html file based on your csv file containing the moving averages. In order to generate the html file your csv file needs to contain the headings year, 10-year_MA as well as data. Example:

$ weatherComparator -v -r "San Francisco" -r Milan -c moving-average-weather-data.csv -d https://raw.githubusercontent.com/amxyz-cyber/data/main/moving-average-weather-data2.csv

Hint: You can leave out the regions but then the default values 'region1' and 'region2' will be used instead.

Running the program:

This repositary currently contains binaries for the Raspberry Pi aarch64 architecture as well as ia64. You will also need the haskell-platform which basically includes the ghc-compiler and its dependencies.

weatherComparator or ./weatherComparator

Install the project

If you want to compile the programm by yourself then please download stack.

After installing stack and having downloaded this project go to the directory containing the project weather-project and enter the following commands:

$ stack build
If Stack complains about that it can't find the lts build or any link provided in packages.yaml is invalid then you need to upgrade stack first:
$ stack upgrade
After compiling the project you can run the program by entering the following command:

$ stack exec weather-project-exe

Note that if you can't use the stack command globally, i.e. you can use stack only locally ./stack then you can export the path to the folder containing its binary, for example:

$ export PATH="${PATH}:/home/username/.local/bin"

The compiled binary will be placed in a subfolder of this project depending on the architecture. For example, for 'aarch64' you'll find the binary in a path similar to this:
.stack-work/dist/aarch64-linux/Cabal-3.2.1.0/build/weather-project-exe/weather-project-exe

To finally install the binary, you'll use the following command:
$ stack install

This command copies the binary to your local bin folder such as
/home/username/.local/bin
You can rename the binary by simply using the move command:

$ mv weather-project-exe weatherComparator

This will work, too.

amxyz-cyber / weather-project