11 Common Issues

Here are some common issues that we’ve seen during consultations. While we’ve covered some of these in prior chapters, it’s worth having a list to review.

11.1 Difference between `=` and `==`

The former, = is used to set something as equal as in x = 5 where the variable x is equal to, or has a value of, 5. Conversely, the double equal == is used to test for equality. For instance, if we set a variable x to equal the value 5, the code x == 5 would return True and the code x == 6 would return False.

11.2 Spelling and capitalization matter

Programming languages are generally case-sensitive. The variable x is different from the variable X. Likewise, a function mean() and a function Mean() would be separate functions.

It’s easier to list the languages which are not case sensitive than those which are:

SQL commands are not case-sensitive, but it’s common practice to see the commands as all-caps, such as SELECT, because it makes reading the argument easier
Functions in Excel are not case-sensitive
HTML tags are generally not case-sensitive, however, class names in HTML and CSS are case-sensitive.

Again, it’s easier to just assume everything is case-sensitive.

11.3 Special characters and reserved words

Many programming languages reserve specific words for specific tasks. For example, a function called sum() is fairly common across languages. While you could make your own alternate function called sum(), this may lead to unexpected results when using the sum() function, so it’s better to avoid using the same name.

Similarly, it is best practice to avoid naming variables or objects the same as common function names. Again, while you could write code like sum = sum(), this will get confusing for you, and may lead to unexpected results and code behavior downstream.

The same is true for certain characters. Many programming languages treat NA as a special class of missing value. This is not the same as "NA", which would instead be a character string containing the letters NA.

11.4 Ending a code chunk

Certain programming languages require a character at the end of a line or statement.

C/C++, Java, and PHP all require a semicolon ; at the end of lines or statements.

Why do we say statement? Because technically you could write your code

if(x == y){cout << "values are equal";x++;cout<<"x was increased & is now 
equal to "<<fixed<<setprecision(1)<<x<<endl;}else if(x > y){cout << "x greater 
than y";x--cout<<"x was reduced & is now equal to "<< fixed <<setprecision(1)
<<x<<endl;} else {cout << "x must be less than y";x++;cout<<"x was increased 
& is now equal to "<<fixed<<setprecision(1)<<x<<endl;}

But, there’s an error in that code. A required semicolon is missing. Can you find it?

Having each statement on its own line is easier to read.

if (x == y) {
  cout << "values are equal";
  x++;
  cout << "x was increased & is now equal to " << fixed << setprecision(1) << x << endl;
} else if (x > y) {
  cout << "x greater than y";
  x--
  cout << "x was reduced & is now equal to " << fixed << setprecision(1) << x << endl;
} else {
  cout << "x must be less than y";
  x++;
  cout << "x was increased & is now equal to " << fixed << setprecision(1) << x << endl;
}

11.5 Order of Execution

In the chapter about functions, we gave an example where libraries which have functions with the same name will default to the last loaded library when called, so it’s good habit to explicitly specify the library when using functions which might have the same name in a different library

load(library_B)
load(library_A)

values <- c(2, 4, 7, 5, 9, NA)

library_B:mean(values) 
#result will be `NA`

library_A:mean(values) 
#result will be `5.4`

11.6 Order of Assignment

I’ve had researchers execute code chunks which assign a value to a variable, then reassigns a different value to the same variable, undoing the first assignment.

researchVariable <- someFunction(data)
researchVariable <- aDifferentFunction(data)
plot_or_graph <- anotherFunction(researchVariable)
all of
the plot
or graph
code sometimes it's a lot

If you see this, ask the researcher to talk through their code, and gently point out the reassignment. If it’s for different conditions, ask if others looking at the code would understand that.

But, often they’ll share that they were trying different things for working with their data. If the assignment is unneeded, they should just remove it from their code. If they are using version control (e.g. git), they will be able to recover the code by looking at previous versions.

But, including code that’s not used in the final script clutters the script, and makes it more difficult to trace when errors occur.

11.7 Matching Delimiters: Quotes, Parens, Brackets

Delimiters are used to indicate the boundaries for something specific in the syntax. They need to be a matching set, so the computer knows the beginning and end boundary. A missing end quote or bracket will often throw an error; review the error trace to find where to complete the pair.

Many text editors (such as Sublime Text or Notepad++) or development environments (such as Jupyter Notebooks or RStudio) will auto-complete the end delimiter, but sometimes you need to adjust that in the options.

Quotes " " or ' ' - used for text. If used within text, you either have to escape it (often the escape character is \), or use the other ("you'll need to surround contractions with double-quotes")
Parentheses ( ) - used for functions or order of operation in expressions
Curly-Brackets { } - often used to indicate code chunks
Square-Brackets [ ] - used to indicate an index in a collection

11.8 Reading and Writing Data Files

Reading and writing files can be a pain point. There are two things to consider when reading from or writing to files

The location of the file with respect to the script that is doing the request.
The type of file to be read/written. Different file types require different functions.

It is useful to understand how to navigate between directories on a command line, as many of the keywords and shortcuts are used to create the path to the files. There are recommended lessons for learning the Bash command line in the Resources chapter.

Here is an example loading a CSV file in Python

import pandas as pd

# updates the event list based on previously created csv
events_df = pd.read_csv("../data/events.csv")

Here is the folder structure. Folders have a / at the end of their name, while files will have an extension (.csv, .r, etc)

- data/
  - events.csv
- src/
  - myPythonFile.ipynb

Here is an example of writing a CSV file in R

write.csv(mydata, file = "output/census_names2.csv", row.names = FALSE)

The folder structure is

- project folder/
  - get_names.r
  - output/
    - census_names2.csv

Here is an example of saving an HTML file (which is just a text file) in Python

import requests
from bs4 import BeautifulSoup

year = "1995"
save_prefix = "ipeds_" 
url = "https://nces.ed.gov/ipeds/datacenter/DataFiles.aspx?year="+year

output_path = "./"+year+"/"+save_prefix+year+"_"+".html"
webpage = requests.get(url)

with open(output_path, 'wb') as file:
  file.write(webpage.content)