Best Linkedin posts about Python

If you need to populate multiple Powerpoint presentations with different data, I cannot recommend highly enough the python-pptx package, which allows the easy editing of the XML code inside Powerpoint documents. You can start with a template file, and then code functions to replace data and text with new data and text. I recently wrote code like this in python and then imported it into a Shiny app where you could select your parameters and click to instantly download your new Powerpoint. Saved hundreds of hours of manual data entry. Highly recommended. #python #rstats #analytics #peopleanalytics #data #datascience

HIRING FOR BACKEND ENGINEER Nicolas Carmont and I are hiring for a backend engineer at our ClimateTech startup. :) We are launching officially soon. Backed by Entrepreneur First, our mission is to democratise access to sustainability data. We are looking for an individual that would like to join as our first technical hire and have a large impact on our business from day one. Tech stack: node.js, Python, AWS Location: London, remote possible too If you know anyone that is interested, please tell them to add me on LinkedIn with a short note saying that they're interested in the job. And yes: Equity options included :) #python #hiring #aws #nodejs

4 Types of Analytics There are 4 types of analytics and you should ideally follow them in the following order. 1. Descriptive: It tells you what happened in the past. You get the data from the past and report what happened with descriptive statistics. Example: we plot the last 12 months' sales, showing sales numbers are going down. 2. Diagnostic: It tells you why something happened in the past. You need to look deeper, build hypotheses, and analyse data to identify reasons behind the patterns found in the descriptive stage. Example: why are sales numbers going down? 3. Predictive: It tells you what will happen in the future. Based on past data and possible future factors you have to predict the future. Machine learning has great applications here. Example: what will be the sales in the next month? 4. Prescriptive: It tells you what to do achieve the desired outcome. You should study cause and effect relationships between independent and dependent variables to recommend what should be the ideal course of action. Example: what to do to increase sales by 30% in 3 months? Which type of Analytics is your favourite and why? 3 best websites for learning Data Science: 1. Coursera: bit.ly/coursera1234 2. DataCamp: bit.ly/datacamp123 3. Udemy: bit.ly/udemy1a SHARE and TAG others so that others can also learn. #DataScienceWithDrAngshu #DataScience #Analytics #BigData #MachineLearning #ArtificialIntelligence #Data #DataAnalytics #Python #SQL #Statistics #DataVisualisation #Interview #Job Follow Dr. AngShuMan Ghosh (Ph.D., MBA, BE) YouTube: youtube.com/drangshu

Evaluating productivity based on lines of code is meaningless, if not downright insulting. Developers and other related professionals use their intellect to create complex systems and solve challenging problems. More often than not, improving the efficiency of your code will result in fewer lines rather than the opposite. We are not typists, but problem solvers that happen to type! Do you agree with this? Let me know your thoughts in the comments. #python #datascience #programming #linkedin #productivity #career

Python Pandas is slow at manipulating large data sets & Deep Learning wouldn't have taken off without Nvidia's awesome GPU's. What if we try to reduce Pandas bottleneck using GPU's? All Thanks to RAPIDS GPU-Accelerated Dataframe Library called cuDF we can achieve great speedup for Pandas like operations. In this video, I'll show you the speed up provided by cudF in Google Colaboratory. I hope you all like it 🙂 Link in the first comment of the post 🙂 #cudf #datascience #machinelearning #pandas #python

⚡️ Some of the most powerful skills in data science: . . . 🗣 1. Communication ❤️ 2. Empathy 🔎 3. Curiosity 👉🏽 The technical skills get you in, but the soft skills is what helps you drive value. #OnTheMarkData #data #datascience #dataengineer #dataanalyst #analytics #ml #machinelearning #python #r #sql ----- 🔔 Want more content like this in your LinkedIn feed? Then don’t forget to click “Follow” on my profile!

Tuning your machine learning models wrongly? This is something which is a common phenomenon with newcomers coming into the field (me included). For some reason I see them trying to tune each and every possible hyperparameter without understanding its meaning. A few tips to make it effective? 1. Don't tune things like random_state, that is simply horrible 2. Read the documentation and understand hyperparameters before using them, including their default values (example: 'auto' and 'sqrt' mean the same thing for max_features in some ensemble models) 3. Bayesian tuning methods can often be a better option vs. vanilla grid search especially when data volume or model complexity is high 4. Some hyperparameters are related, so tuning one might affect the other (example: max_depth is a cruder way to restrict tree depth but you also have min_samples_split and min_samples_leaf for more fine-grained control. Don't set conflicting values in the search space, some will never even affect the model but you lose valuable computation time) 5. Default model hyperparameters are good in most cases, tuning is not a magic wand that you will always get 5-10% performance improvement Hope this helps as you employ hyperparameter tuning in your experiments. Feel free to share your own tips in the comments! #machinelearning #data #analytics #datascience #deeplearning #python #statistics #artificialintelligence #ai

Since the last 3-5 years we are hearing Python/R will be dead by the next year, long live 'X' language. Data scientists will be obsolete in the next few years and will be replaced by software engineers and Auto-ML. Big data and spark are dead and Kubernetes is the future. Edit: I forgot the icing on the cake. SQL is dead and NoSQL is what every enterprise needs. cc Eric Weber ;) Still waiting to see when that 'next year' is coming. Till then we all have to enjoy the clickbait articles and webinars I guess. 🙂 #datascience #machinelearning #deeplearning #ai #bigdata #artificialintelligence #python