I Quit Coding - How I use ChatGPT instead as Data Analyst

[@CodingIsFun]

*Hey guys! I'm curious about your thoughts on the rapid development of AI. Are you already using it in your day-to-day workflows, or do you have concerns about its impact on our industry? Let's discuss in the comments below!* ChatGPT Excel add-in ⤵ 👉 pythonandvba.com/mytoolbelt Source code (generated by ChatGPT) ⤵ 📑 github.com/Sven-Bo/Practical-Examples-for-Using-ChatGPT-in-Data-Analysis Links to the ChatGPT conversation history ⤵ 👉 Excel Report: sharegpt.com/c/61O3GXq 👉 Refactor Code & PowerQuery Steps: shareg.pt/mrVjAWd 👉 SQL Generation: shareg.pt/1fWlHku

[@keeplearning9897]

AI has become a topic of discussion for everyone, with people on online platforms explaining its capabilities. However, it is important to realize that AI applications have already started to become widespread and could potentially consume our minds. Soon, people may depend entirely on AI, as machines are now capable of learning. Rather than solely discussing the benefits of AI. @CodingIsFun: You are good content creator, but your last videos (6-8 videos) are talking about the solution using AI only, So it will be eat human mind gradually. It would be beneficial to also explore ways in which we can outsmart and overcome it.

[@chamamemestre]

Thank you. Better, fast, and non bloated explanation like I see in many paid courses. Good job!

[@CodingIsFun]

Glad you liked it. Thanks for watching. ♥

[@michaelharper7029]

I am just starting to explore how I can use ChatGPT in my work as a database specialist. I learned SO much from this video. It was also very easy to watch and no time wasted. Thank you.

[@CodingIsFun]

Happy to hear that it was useful; thank you for taking the time to leave a comment and for watching the video! 👍

[@MyCodingDiary]

Your explanations are so concise and easy to follow😀. Thank you for making this video!

[@CodingIsFun]

Thanks for leaving a comment and for taking the time to watch! Glad you liked it. 🤩

[@torque6389]

Great video as always! Please keep the great content coming.

[@CodingIsFun]

Thanks for the positive feedback! Appreciate you taking the time to leave a comment. 😃

[@tuandhtl]

great video, thanks Sven-Bo

[@CodingIsFun]

My pleasure! Appreciate you taking the time to watch and leave a comment. 👍

[@xilllllix]

great vid! this shows how powerful chatgpt can really be in the right hands...

[@CodingIsFun]

Thanks for leaving a comment and for taking the time to watch! Glad you liked it. 👍

[@bc4198]

Fantastic video! I am the only one in my office aware of ChatGPT/AI at all, and plan to be the robots' best friend when they obsolesce the other jobs 😅... I use ChatGPT/AI for myself frequently, but little for work so far. I've snuck it into a couple little test cases and have some long-term ideas, but will have to train it on custom data before I can really use it day-to-day.

[@CodingIsFun]

Thanks for watching! Haha, I'm right there with you on being the robots' best friend - I even use "please" in my prompts, just in case 😅

[@bc4198]

​@@CodingIsFun Yay! I sensed that you did, but wasn't sure whether to ask 😊

[@pleabargain]

Thanks for posting. You have a subscription which gives you larger number of tokens, correct?

[@CodingIsFun]

Thanks for watching. I do have the ChatGPT Plus subscription, but as mentioned in the video (2:22 min), for all examples I have used the free GPT-3.5 model.

[@rajeevgangal542]

Does the Plus subscription make any difference? If yes what? Initially the free version used to generate complete python cose however complex. Nowadays it often stops midway and repeat regenerate and task submission are needed to get complete code.

[@CodingIsFun]

Thanks for watching and asking! I haven't really compared the GPT-3.5 and GPT-4 models in terms of code generation. But, in general, the GPT-4 model is usually better with longer prompts and conversations, and it makes fewer mistakes too.

[@jhari88]

Sorry I cant find it on the discord or website but what Python IDE are you using? I'm just starting so copying mostly what others are doing.

[@CodingIsFun]

Thanks for watching. I am using VS Code

[@jhari88]

@@CodingIsFun Perfect! Thanks! keep up the good work!

[@brilliancetventertainment9416]

hello Coding Is Fun.....do you have a video on using python to search a excel workbook with several sheets in it for a specific search term in cell 1 of each sheet and once found copy the entire column and move it to a different column?....eg. say the cell 1 text is "PORTS" and it is in all excel workbooks that has several sheets within it , once the string in cell 1 is found and not already in column B then we want to copy that entire column and paste it in column B and then remove all blank cells from the column. Also, just to mention that cell 1 of all the workbooks are merged cells. ChatGPT gave code that didnt work at all, I even created a macro and provided the VBA to it to convert to python which the code decriptions seems to suggest the right thing but it just does work. I think this one if not done already makes for a great video topic

[@CodingIsFun]

Thanks for tuning in! While I don't have a specific video addressing your query, I do have one where I demonstrate how to search and replace text in multiple Excel files. It could be a helpful starting point! Feel free to give it a watch here: kzbin.info/www/bejne/i6raapmgebCMh7s Happy Coding! 🤓

[@brilliancetventertainment9416]

@@CodingIsFun thanks man ! will try it out and let you know

[@brilliancetventertainment9416]

@@CodingIsFun here is the code that chatGPT gave, can you look at this and indicate what is issue is?: import os from openpyxl import load_workbook from openpyxl.utils import column_index_from_string, get_column_letter # Path to directory containing all excel files folder_path = r'C:\Users\user\Downloads\Sheets_B_O' # Search terms in cell 1 olt_card_search_term = 'OLT CARD-PORT' olt_slot_search_term = 'OLT SLOT' fat_search_term = 'FAT' # Loop through each file in the directory for filename in os.listdir(folder_path): if not filename.endswith('.xlsx'): continue # Skip non-excel files # Load the workbook wb = load_workbook(os.path.join(folder_path, filename)) # Initialize a flag to indicate if any sheet contains the search terms found_search_terms = False # Loop through each sheet in the workbook for sheet in wb: # Initialize variables olt_card_col = None olt_slot_col = None # Find the index of the first row that contains the search term for row in sheet.iter_rows(min_row=1, max_row=1): for cell in row: if olt_card_search_term in str(cell.value): olt_card_col = cell.column print(f"Found {olt_card_search_term} in column {get_column_letter(olt_card_col)} of sheet {sheet.title}") elif olt_slot_search_term in str(cell.value): olt_slot_col = cell.column print(f"Found {olt_slot_search_term} in column {get_column_letter(olt_slot_col)} of sheet {sheet.title}") # Check if columns were found if olt_card_col is not None and olt_slot_col is not None: found_search_terms = True # Copy OLT CARD-PORT and OLT SLOT columns to columns A and B, respectively for i in range(1, sheet.max_row + 1): sheet.cell(row=i, column=1).value = sheet.cell(row=i, column=olt_card_col).value sheet.cell(row=i, column=2).value = sheet.cell(row=i, column=olt_slot_col).value # Copy FAT column to column C for column in sheet.columns: cell_1 = column[0] if str(cell_1.value) == fat_search_term: found_search_terms = True for i in range(1, sheet.max_row + 1): sheet.cell(row=i, column=3).value = column[i - 1].value print(f"Found {fat_search_term} in column C of sheet {sheet.title}") break # Stop searching for FAT once found # Delete all other columns except A, B and C for column in reversed(range(4, sheet.max_column + 1)): for cell in sheet[column]: cell.value = None # Delete all other columns except A, B and C for column in reversed(range(4, sheet.max_column + 1)): if any(column in merged for merged in sheet.merged_cells.ranges): continue # Skip merged cells for cell in sheet[column]: cell.value = None # Rename the workbook if any of the sheets contain the search terms if found_search_terms: new_filename = f"NEW_{filename}" os.rename(os.path.join(folder_path, filename), os.path.join(folder_path, new_filename)) print(f"Workbook renamed to '{new_filename}'") # Save the modified workbook if found_search_terms: # Create a new file name by appending 'MODIFIED' to the original file name new_filename = os.path.splitext(filename)[0] + '_MODIFIED.xlsx' wb.save(os.path.join(folder_path, new_filename)) print(f"Workbook saved as {new_filename}") else: wb.save(os.path.join(folder_path, filename)) print("Workbook saved without modification") error in python IDE: C:\Users\user\PycharmProjects\pythonProject\OLTExcelparsing\venv\Scripts\python.exe C:/Users/user/PycharmProjects/pythonProject/OLTExcelparsing/venv/excelparsing4OLT.py Found OLT CARD-PORT in column D of sheet JCA00209 Traceback (most recent call last): File "C:\Users\user\PycharmProjects\pythonProject\OLTExcelparsing\venv\excelparsing4OLT.py", line 65, in if any(column in merged for merged in sheet.merged_cells.ranges): File "C:\Users\user\PycharmProjects\pythonProject\OLTExcelparsing\venv\excelparsing4OLT.py", line 65, in if any(column in merged for merged in sheet.merged_cells.ranges): File "C:\Users\user\PycharmProjects\pythonProject\OLTExcelparsing\venv\lib\site-packages\openpyxl\worksheet\merge.py", line 137, in __contains__ return coord in CellRange(self.coord) File "C:\Users\user\PycharmProjects\pythonProject\OLTExcelparsing\venv\lib\site-packages\openpyxl\worksheet\cell_range.py", line 263, in __contains__ cr = self.__class__(coord) File "C:\Users\user\PycharmProjects\pythonProject\OLTExcelparsing\venv\lib\site-packages\openpyxl\worksheet\cell_range.py", line 51, in __init__ if "!" in range_string: TypeError: argument of type 'int' is not iterable Process finished with exit code 1

[@VayunEkbote]

Hey great video! What are your thoughts on the ChatGPT code interpreter plugin? Will it put an end to data analyst jobs? I read another comment asking a similar question to which you replied that humans will take the final decision and guide the AI. But will human intervention really be necessary after a certain point? And if yes, then is a highly qualified data analyst required for that purpose, as even a beginner can do the job with AI.

[@brando2818]

Given enough time and progress with ai, probably not if everyone has access to this. I have access to a lot of plugins and am a data analyst, and I can see a future where many are not needed. I don't think this would happen overnight though.. I could do everything in this video with code interpreter in less than 5 mins. If this tool(code interpreter) or something similar was connected to a datasource, an end user could just prompt for whatever data question they had I mind.. furthermore, at some point, the system could deliver this information unprompted.

[@VayunEkbote]

@@brando2818 as a data analyst, do you think this job will even exist in the future? It is really difficult to predict what jobs will be eradicated by AI, and me being a college student, I am really interested in knowing what I should do to prevent getting caught in this "AI trap"

[@CodingIsFun]

Hey, thanks for watching and your great question! I haven't tried the ChatGPT code interpreter plugin yet, but once I get the chance, I'll definitely check it out and let you know what I think. I've seen a couple of comments asking about how data roles might change with all these AI advancements. Maybe I'll make a video to share my thoughts on that. In a perfect world, if all data were super organized (like in a data warehouse), AI tools would be amazing at analyzing data, creating charts, and stuff like that. But in real life, most data is kind of all over the place, spread across a ton of Excel files. And if you need data from different departments, that's probably even more files to deal with. A lot of my colleagues spend a lot of time just working with Excel to get things ready for analysis. So, considering all this, I think data engineers might become even more important in the future. And at least for the next few years, I believe we'll still need humans to guide AI and present data to the higher-ups. The world of data is always changing, but if you've got a solid foundation in data engineering and analysis and can adapt to new tech, you should be in a good spot to handle whatever comes your way. But that's just my personal opinion.

[@VayunEkbote]

@@CodingIsFun I will be looking forward to the code interpreter video!

[@History_Mystery_Crime]

Data analysts will be there, but the number will decline. Teams consisting of 10 can be easily reduced to 3 with AI

[@franciscozapata7683]

Hi, the company I work for, doesn't allow us to use chat GPT even though I work with excel every single day.

[@CodingIsFun]

That's a pity! :(

[@bwilliams060]

This is a great video. Shows how non coders can leverage python and LLMs to automate safely with limited experience. Thanks!

[@CodingIsFun]

My pleasure! Appreciate you taking the time to watch and leave a comment. 👍

[@steeful]

Question? Would my data now become part of ChatGPT? So if someone knew I was using it could someone else ask questions about MY data?

[@CodingIsFun]

Thanks for watching! In the video, I didn't actually share any real data, just talked about its structure. As for your question, I agree that sharing sensitive company data with ChatGPT isn't a good idea. It's best to keep your data secure and not risk it becoming part of the AI's knowledge.

[@ESPjump]

Actually, I can make it much faster and personalized with coding. ChatGPT is ok to ask doubts, as a Stackoverflow more eficient. But it didn't help me so much with the analysis.

[@CodingIsFun]

Thanks for watching and throwing in your two cents. 👍

[@matt_milack]

Does it makes any sense anymore spending serious amount of time studying for becoming a data analyst, considering the rise of AI, and what it can be in the near future?

[@CodingIsFun]

Hey Jonas Alderson! Thanks for watching and your excellent question. Predicting the future is always a challenge (even for a Data Analyst 😅), and I can't say for certain how Data roles will transform. That said, I believe humans will continue to make the final decisions and present data to management. As no/low code tools with AI capabilities keep advancing, the tasks of collecting and cleaning data might evolve. As demonstrated in my video, Data Analysts could end up being more like "directors" of data, focusing less on coding and more on guiding the tools. So, while the future's uncertain, being adaptable and embracing new tech will always be a plus in the data analysis world. At the end of the day, it's still up to you whether you want to dive into the data analyst world and learn the existing tools and new technologies. Just remember, this is only my humble opinion.

[@bc4198]

Ask ChatGPT this question 😂. For real though - yes, it's worthwhile, but do it with AI! There has to be humans in charge of knowing what they want AI to do, and you can be the first gen of analyst that's a native user, and come out stronger through the transition to an AI-driven economy. Most people and companies are not even close to ready for, or even aware of AI, and they need you to know all that and more. Put another way, even if AI could do the analysis, your bosses and customers still don't know what to do with the results. Anything you can learn can be valuable.

[@WINCLUBdotCA]

Anyone can tell me if that would be the best setup to analyze Roulette spins made by a specific table? Also I'm trying to build the perfect prompt to get GPT to tell me the next number (within 6 chips) using that info, like an algorithm would. I've seen a few built that gives the exact number, but they won't share or sell.

[@hamidreza1998]

Bro your videos should be illegal 😂😂❤❤ thank you for helping big time

[@CodingIsFun]

Glad I could help! Thanks for watching and for the kind words. ♥

[@leandrov07013]

So I shouldn't go deep on Python, if I want to become a data analyst?

[@yarpenzigrin1893]

No need. You don't need to know object oriented programming for data analysis.

[@CodingIsFun]

In my opinion, it really depends on the specific requirements of the Data Analyst role you're aiming for. While I am not a software developer myself and don't consider myself an expert in Python, I do think it's worthwhile to learn the basics of Python, especially popular plotting and data wrangling libraries, in order to be proficient in data analysis. As I mentioned in the video at the 11:01 mark, understanding the code generated by ChatGPT is important in order to ensure that you don't accidentally harm your computer or data. Additionally, having a good grasp of Python can help you to create better prompts and leverage AI tools to your advantage.

[@leandrov07013]

@@CodingIsFun excellent, thanks for both the great content and this explanation

[@bc4198]

It's not a prerequisite to applying, but I think you'd only feel regret if you were at a job that won't let you have the right tools (many corp & gov jobs). If you can use Python, you should learn it, if only for scripting / automating your own tasks. Excel and SQL are bigger deal breaker skills IMO, where you can't do jobs without them; no matter what tool you use for analysis, you have to be able to get to the data to begin with, however it is stored.

[@leandrov07013]

Thank you guys

[@swiatlowiekuiste]

I find risky making available to chat gpt real data. I'd recommend using made up templates instead

[@CodingIsFun]

Thanks for watching. Totally agree, it's not smart to share sensitive data with ChatGPT. Better to play it safe and keep your data secure, so it doesn't end up as part of the AI's knowledge.

[@swiatlowiekuiste]

@@CodingIsFun Yes, not to mention your boss wouldn't be happy if you shared the company's secrets :)

[@EveryoneStoreOnline]

how to create excel convert exe file

[@CodingIsFun]

I don't know what you mean