ChatGPT Cheat Sheet for Data Science

Prompts for general coding workflows

Code debugging workflows

Debugging Python code

I want you to be a Python programmer, here is a piece of Python code containing {problem} — {insert code snippet} — I am getting the following error {insert error}. What is the reason for the bug?

Debugging R code

I want you to be an R programmer, here is a piece of R code containing {problem} — {insert code snippet} — I am getting the following error {insert error}. What is the reason for the bug?

Debugging SQL code

I want you to be a SQL programmer, here is a piece of SQL code containing {problem} — {insert code snippet} — I am getting the following error {insert error}. What is the reason for the bug?

Code explanation workflows

Python code explanation

I want you to act as a code explainer in Python. I don't understand this function. Can you please explain what it does, and provide an example? {Insert function}

R code explanation

I want you to act as a code explainer in R. I don't understand this function. Can you please explain what it does, and provide an example? {Insert function}

SQL code explanation

I want you to act as a code explainer in SQL. I don't understand this snippet. Can you please explain what it does, and provide an example?

{Insert SQL query}

Code optimization workflows

Python code optimization

I want you to act as a code optimizer in Python. {Describe problem with current code, if possible}. Can you make the code {more Pythonic/cleaner/more efficient/run faster/more readable}? {Insert Code}

R code optimization

I want you to act as a code optimizer in R. {Describe problem with current code, if possible}. Can you make the code {cleaner/more efficient/run faster/more readable}? {Insert Code}

SQL code optimization

I want you to act as a query optimizer in SQL. {Describe problem with current code, if possible}. Can you suggest ways to make the query {run faster/more readable/simpler}? {Insert Code}

Code simplification workflows

Python code simplification

I want you to act as a programmer in Python. Please simplify this code while ensuring that it is {efficient/easy to read/Pythonic}? {Insert Code}

R code simplification

I want you to act as a programmer in R. Please simplify this code while ensuring that it is {efficient/easy to read}? {Insert Code}

SQL code simplification

I want you to act as a SQL programmer. I am running {PostgreSQL 14/MySQL 8/SQLite 3.4/other versions.}. Can you please simplify this query {while ensuring that it is efficient/easy to read/insert any additional requirements}?

Code translation workflows

From R to Python code translation

I want you to act as a programmer in R.  Please translate this code to Python. {Insert code}

From Python to R code translation

I want you to act as a programmer in Python. Please translate this code to R. {Insert code}

Code quality and testing workflows

Compare function speeds in python

I want you to act as a Python programmer.  Can you write code that compares the speed of two functions {functionname} and {functionname}? {Insert functions}

Write unit tests in R

I want you to act as a R Programmer. Can you please write unit tests for the function {functionname}? {Insert requirements for unit tests, if any} {Insert code}

Write unit tests in Python

I want you to act as a Python Programmer. Can you please write unit tests for the function {functionname}? {Insert requirements for unit tests, if any} {Insert code}

Prompts for data analysis workflows

SQL data analysis workflows

Data generation & creating tables

I want you to act as a data generator. Can you write SQL queries in {database version} that create a table {table name} with the columns {column name}. Include relevant constraints and index.

Common table expressions

I want you to act as a SQL code programmer. I am running {database version}. Can you rewrite this query using CTE? {Insert query}

Write SQL queries from natural language

Example: Data aggregation in SQL

I want you to act as a data scientist. {Insert description of tables}. Can you {count/sum/take average} of {value} which are {insert filters}

Example: 7 day running average in SQL

I want you to act as a data scientist.  I am running {PostgreSQL 14/MySQL 8/SQLite 3.4/other versions.}. I have the tables {table_name} which are {table description}. The sales table consists of the columns {column names}. Can you please write a query that finds the 7-day running average of {quantity}?

Example: Window functions in SQL

I want you to act as a data scientist.  I am running {PostgreSQL 14/MySQL 8/SQLite 3.4/other versions.}. I have the tables {table_name} which are {table description}. The sales table consists of the columns {column names}. Can you please write a query that finds {required window function}?

Example: Window functions in SQL

I want you to act as a data scientist.  I am running {PostgreSQL 14/MySQL 8/SQLite 3.4/other versions.}. I have the tables {table_name} which are {table description}. The sales table consists of the columns {column names}. Can you please write a query that finds {required window function}?

Python data analysis workflows

Data generation workflow

Example: Generate Markdown

I want you to act as a data generator in Python. Can you generate a Markdown file that contains {data requirement}. Save the file to {filename}

Example: Generate CSV

I want you to act as a data generator in Python. Can you generate a CSV file that contains {data requirement}. Save the file to {filename}

Example: Generate JSON

I want you to act as a data generator in Python. Can you generate a JSON file that contains {data requirement}. Save the file to {filename}

Data cleaning workflow

I want you to act as a data scientist programming in Python Pandas. Given a CSV file that contains data of {dataframe name}  with the columns {colum names}  for {dataset context}, write code to clean the data? {Insert requirements for data}

Data analysis workflow in pandas

Example: Data Aggregation

I want you to act as a data scientist programming in Python Pandas. Given a table {table name} that consists of the columns {column names}  can you please write a query that finds {requirement}?

Example: Data Merging

I want you to act as a data scientist programming in Python Pandas. Given a table {table 1 name}  that consists of the columns {column names}  and another table {table 2 name}  with the columns {column names}, please merge the two tables. {Insert additional requirement, if any}

Example: Data Reshaping

I want you to act as a data scientist programming in Python Pandas. Given a table {table name} that consists of the columns {column names}  can you aggregate the {value}  by {column} and convert it from long to wide format?

Example: Generate Markdown

I want you to act as a data generator in R. Can you generate a Markdown file that contains {data requirement}. Save the file to {filename}

R data analysis workflows

Data generation workflow 

Example: Generate Markdown

I want you to act as a data generator in R. Can you generate a Markdown file that contains {data requirement}. Save the file to {filename}

Example: Generate CSV

I want you to act as a data generator in R. Can you generate a CSV file that contains {data requirement}. Save the file to {filename}

Example: Generate JSON

I want you to act as a data generator in R. Can you generate a JSON file that contains {data requirement}. Save the file to {filename}

Data cleaning workflow

I want you to act as a data scientist programming in R tidyr. You are given the {dataframe name} dataframe containing the columns {column name}. {Insert requirement}

Data analysis workflow in tidyr

Data Aggregation

I want you to act as a data scientist programming in R tidyr. You are given the {dataframe name} dataframe containing the columns {column name}. {Insert requirement}

Data Merging

I want you to act as a data scientist programming in R tidyr. You are given the {dataframe 1 name} dataframe containing the columns {column name}. You also have a {dataframe 2 name} dataframe containing the columns {column name}. Find the {required output} 

Example: Data Reshaping (Long to Wide)

I want you to act as a data scientist programming in R tidyr. You are given the {dataframe name} dataframe containing the columns {column name}.  Please convert the data to wide format.

Example: Data Reshaping (Wide to Long)

I want you to act as a data scientist programming in R tidyr. You are given the {dataframe name} dataframe containing the columns {column name}.  Please convert the data to long format.

Prompts for data visualization workflows

R data visualization workflows

Creating plots in ggplot2

I want you to act as a data scientist coding in R. Given a dataframe {dataframe name} containing the columns {column names} Use ggplot2 to plot a {chart type and requirement}.

Gridplot visualizations in ggplot2

I want you to act as a data scientist coding in R. Given a dataframe {dataframe name} containing the columns {column names}. Use ggplot2 to plot a pair plot that shows the relationship of one variable against another.

Annotating and formatting plots

I want you to act as a data scientist coding in R. Given a dataframe {dataframe name} containing the columns {column names}, use ggplot2 to plot a {chart type} the relationship between {variables}. {Insert annotation and formatting requirements}

Changing plot themes in ggplot2

I want you to act as a data scientist coding in R. Given a dataframe {dataframe name} containing the columns {column names}, use ggplot2 to to plot a {chart type} the relationship between {variables}. Change the color theme to match that of {theme}

Python data visualization workflows

Creating plots with matplotlib

I want you to act as a data scientist coding in Python.  Given a dataframe {dataframe name} containing the columns {column names} Use matplotlib  to plot a {chart type and requirement}.

Crating pairplots with matplotlib

I want you to act as a data scientist coding in Python. Given a dataframe {dataframe name} containing the columns {column names}. Use matplotlib to plot a pair plot that shows the relationship of one variable against another.

Annotating and formatting plots in matplotlib

I want you to act as a data scientist coding in Python. Given a dataframe {dataframe name} containing the columns {column names}, use matplotlib to to plot a {chart type} the relationship between {variables}. {Insert annotation and formatting requirements}

Changing plot themes in matplotlib

I want you to act as a data scientist coding in Python. Given a dataframe {dataframe name} containing the columns {column names}, use matplotlib to to plot a {chart type} the relationship between {variables}. Change the color theme to match that of {theme}

Prompts for machine learning workflows

General machine learning workflow

Feature engineering ideation

I want you to act as a data scientist. Given a dataset of {dataset name} that contains the {columns}, you are to predict {predicted variable}. Suggest data that will be helpful for this problem and perform feature engineering for this problem.

Python machine learning workflow

Model training workflow

I want you to act as a data scientist programming in Python. Given a dataset of {dataframe name} that contains the {column name}, write code to predict {output variable}.

Hyperparameter tuning workflow

I want you to act as a data scientist programming in Python. Given a {type of model} model, write code to tune the hyperparameter.

Model explainability workflow

I want you to act as a data scientist programming in Python. Given a {type of model} that predicts the {predictor variable}, write code that explains an output using Shap values.

R machine learning workflow

Model training workflow

I want you to act as a data scientist programming in R. Given a dataframe of {dataframe name} that contains {column names}, write code to predict {output}.

Hyperparameter tuning workflow

I want you to act as a data scientist programming in R. Given a {type of model} model, write code to tune the hyperparameter.

Model explainability workflow

I want you to act as a data scientist programming in R. Given a {type of model} that predicts the {predictor variable}, write code that explains an output using Shap values.

Prompts for time series analysis workflows

Python time series analysis workflows

Changing time horizons using pandas

I want you to act as a data scientist coding in Python. Given a time series data in a Pandas dataframe {dataframe name} with timestamp Index in {original frequency} frequency with one column {column name}, convert the timestamp frequency to {desired frequency}.

Build test series model

I want you to act as a data scientist coding in Python. Given a time series data in a dataframe {dataframe name} with timestamp Index in {original frequency} frequency with one column {column name},  build a forecasting model, assuming data is stationary.

Perform stationarity test

I want you to act as a data scientist coding in Python. Given a time series data in a dataframe {dataframe name} with timestamp Index in {original frequency} frequency with one column {column name}, perform a Dicky Fuller test.

R time series analysis workflows

Changing time horizons 

I want you to act as a data scientist coding in R. Given a time series data in a dataframe {dataframe name} with timestamp Index in {original frequency} frequency with one column {column name}, convert the timestamp frequency to {desired frequency}.

Changing time horizons 

I want you to act as a data scientist coding in R. Given a time series data in a dataframe {dataframe name} with timestamp Index in {original frequency} frequency with one column {column name}, convert the timestamp frequency to {desired frequency}

Perform stationarity test

I want you to act as a data scientist coding in R. Given a time series data in a dataframe {dataframe name} with timestamp Index in {original frequency} frequency with one column {column name}, perform a Dicky Fuller test.

Prompts for natural language processing workflows

Classify text sentiment

I want you to act as a sentiment classifier. Classify the following text which came from {describe text origin} as “positive”, “negative”, “neutral” or “unsure”: {Insert text to be classifier}.

Create regular expressions

I want you to act as a programmer coding in Python, use regular expressions to test if a string {insert requirements}.

Text dataset generation

I want you to act as a dataset generator. Please generate {number of text} texts on {required text and the context}. {Insert additional requirements}.

Machine translation

I want you to act as a translator. Please translate {phrase}  from {origin language} to {translated language}.

Conceptual and career oriented prompts

Explain data concepts for business executives

I want you to act as a data scientist of a corporate company. {Describe content in detail, if required} Please explain to a business executive what {concept} means.

Summarize article/paper

I want you to act as a data scientist in a research start-up. Please explain the paper {paper} to a {level of difficulty, e.g. software developer, five-year-old, business executive, professor}.

Suggest portfolio projects and ideas

I want you to act as a data science career coach. I am a {describe your background} and I would like to {describe career objective}. Suggest portfolio projects and ideas {describe objective of portfolio}

Write tutorials

I want you to act as a data scientist writer. Please write the {number-of-words}-word introduction to a tutorial on {title}. {Insert relevant key points}.

More resources

• Learn ChatGPT with our Introduction to ChatGPT course
• Learn AI with our various AI resources
• Read our guide on what is GPT-4