On Christmas 2023, to clear my mind, I decided to take a break from Python and R and try something new: I completed the Introduction to Julia course in Datacamp.

The course is really good and easy to follow for an absolute beginner like I was, but all the exercises are run within the perfect enviroment of the course provider. It is later, when you try on your own PC, when the problems arise.

First you have to install the Julia language in your computer (instructions in the Julia website). After that, you can install any packages you may need by pressing the key ] when in the Julia REPL and then writing add “whatever” . For example, if you want to write your Julia code in Jupyter notebooks; you will need to install the IJulia pakage.

For me, with all the data analysis languages and tools, the trickiest part is loading the data into the working space. So, after finding out a way to do it, I saved the code here for future reference. Feel free to copy & use it.

JuliaExercise1

Julia lessons for Inma in Jupyter¶

How to retrieve a .csv data file from the web.

Imagine you want to analyse the dataset from the Public Health Scotland website: https://www.opendata.nhs.scot/dataset/c4db1692-fa02-4a1c-af4c-6039c74633ea/resource/29452b1f-a7be-4e93-9e22-dfa120c2df26 (contains data on alcohol related hospital statistics by sex and age group).

For a reproducible analysis, the best way is to retrieve the data directly from the web without downloading the file into your computer (Althought the data could be updated by the data owner and the subsecuents analysis will differ slightly).

Note: for Jupyter to work with Julia code you first need to install IJulia in your Julia REPL

In [21]:

import DataFrames, CSV, HTTP

# read csv file
path = ("https://www.opendata.nhs.scot/dataset/c4db1692-fa02-4a1c-af4c-6039c74633ea/resource/29452b1f-a7be-4e93-9e22-dfa120c2df26/download/arhs_agegender_28_02_2023.csv")
data = HTTP.get(path).body
csv_file = CSV.File(data)
df = DataFrames.DataFrame(csv_file)
    
# Select the columns we are interested in 
alcohol_stays = DataFrames.select(df, :Condition, :FinancialYear, :Gender, :AgeGroup, :NumberOfStays)

# select data from 2021/22
alcohol_stays_2021 = DataFrames.filter(row -> row.FinancialYear == "2021/22", alcohol_stays)

# print the first two rows
println(first(alcohol_stays_2021, 2))

2×5 DataFrame
 Row │ Condition               FinancialYear  Gender   AgeGroup  NumberOfStays 
     │ String                  String7        String7  String31  Int64         
─────┼─────────────────────────────────────────────────────────────────────────
   1 │ All alcohol conditions  2021/22        Male     All               22728
   2 │ All alcohol conditions  2021/22        Female   All               10332

In [ ]:

Calling packages: If for calling the packages you prefer to use using instead of import you don't need to write DataFrames. before every function from the pkg DataFrames, but you still need to write CSV. and HTTP.

In [28]:

using DataFrames, CSV, HTTP

# read csv file
path = ("https://www.opendata.nhs.scot/dataset/c4db1692-fa02-4a1c-af4c-6039c74633ea/resource/29452b1f-a7be-4e93-9e22-dfa120c2df26/download/arhs_agegender_28_02_2023.csv")
data = HTTP.get(path).body
csv_file = CSV.File(data)
df = DataFrame(csv_file)
    
# Select the columns we are interested in 
alcohol_stays = select(df, :Condition, :FinancialYear, :Gender, :AgeGroup, :NumberOfStays)
println(first(alcohol_stays,10))

#select data from 2020/21
alcohol_stays_2019 = filter(row -> row.FinancialYear == "2019/20", alcohol_stays)
println(first(alcohol_stays_2019,2))

10×5 DataFrame
 Row │ Condition               FinancialYear  Gender   AgeGroup  NumberOfStays 
     │ String                  String7        String7  String31  Int64         
─────┼─────────────────────────────────────────────────────────────────────────
   1 │ All alcohol conditions  1997/98        Male     All               21462
   2 │ All alcohol conditions  1997/98        Female   All                8232
   3 │ All alcohol conditions  1998/99        Male     All               21930
   4 │ All alcohol conditions  1998/99        Female   All                8637
   5 │ All alcohol conditions  1999/00        Male     All               23637
   6 │ All alcohol conditions  1999/00        Female   All                8955
   7 │ All alcohol conditions  2000/01        Male     All               23337
   8 │ All alcohol conditions  2000/01        Female   All                9051
   9 │ All alcohol conditions  2001/02        Male     All               24474
  10 │ All alcohol conditions  2001/02        Female   All                9846
2×5 DataFrame
 Row │ Condition               FinancialYear  Gender   AgeGroup  NumberOfStays 
     │ String                  String7        String7  String31  Int64         
─────┼─────────────────────────────────────────────────────────────────────────
   1 │ All alcohol conditions  2019/20        Male     All               25203
   2 │ All alcohol conditions  2019/20        Female   All               11340

www.inmaruiz.com

In [ ]:

Julia Language

Julia lessons for Inma in Jupyter¶

Sponsored content