Posts

Showing posts with the label Big data Server

Featured Post

Top Questions People Ask About Pandas, NumPy, Matplotlib & Scikit-learn — Answered!

Image
 Whether you're a beginner or brushing up on your skills, these are the real-world questions Python learners ask most about key libraries in data science. Let’s dive in! 🐍 🐼 Pandas: Data Manipulation Made Easy 1. How do I handle missing data in a DataFrame? df.fillna( 0 ) # Replace NaNs with 0 df.dropna() # Remove rows with NaNs df.isna(). sum () # Count missing values per column 2. How can I merge or join two DataFrames? pd.merge(df1, df2, on= 'id' , how= 'inner' ) # inner, left, right, outer 3. What is the difference between loc[] and iloc[] ? loc[] uses labels (e.g., column names) iloc[] uses integer positions df.loc[ 0 , 'name' ] # label-based df.iloc[ 0 , 1 ] # index-based 4. How do I group data and perform aggregation? df.groupby( 'category' )[ 'sales' ]. sum () 5. How can I convert a column to datetime format? df[ 'date' ] = pd.to_datetime(df[ 'date' ]) ...

Real thoughts on IBM power8 servers to use on analytics

Image
IBM Servers International Business Machines Corp, in its latest attempt at reviving demand for its hardware products, is launching high-end system servers that it says are 50 times faster than its closest competitor at analysing data.  The POWER8 servers , the product of a $2.4 billion, three-year investment, are part of the company's decade-long shift to higher-value hardware technology.    IBM  said the machines are 50 times faster than the low-end x86-based servers it sold to Chinese PC maker  Lenovo  Group Ltd in January.  The technology services provider said on Wednesday it hopes the servers, designed for large-scale computing, will appeal to clients looking to manage new types of social and mobile computing and mass amounts of data. Last week, the company reported its lowest quarterly revenue in five years, weighed down by falling demand for its storage and server products. IBM dominates the higher-end server market with 57 percent ...