Why scikit numpy and pandas

scikit has machine learning algs.
Scikit needs nd arrays Numpy datatypes
Pandas provide dataframe and R like syntax...

so read data..convert to pandas...numpy..leverage scikit

https://www.datascienceretreat.com/courses/numpy-scipy-pandas-scikit-learn?locale=en

Numpy and Scipy took python from a general programming language to a very powerful matrix-oriented one. Pandas brought data.frames to python. Data.frames are one of the core concepts in modern data analysis. Building on top of these data structures, Scikit-learn brought killer implementations of best-of-breed algorithms, all under a standardized library. Nowadays, python is the programming language of choice of data scientists.
Preprocessing with Pandas

    Reading data
    Selecting columns and rows
    Filtering
    Vectorized string operations
    Missing values
    Handling time
    Time series

NumPy, SciPy

    Arrays
    Indexing, Slicing, and Iterating
    Reshaping
    Shallow vs deep copy
    Broadcasting
    Indexing (advanced)
    Matrices
    Matrix decompositions

Scikit-learn

    Feature extraction
    Classification
    Regression
    Clustering
    Dimension reduction
    Model selection




Comments

Post a Comment

Popular posts from this blog

ScoreCard Model using R

Example R Notebook for ScoreCard Model One often receives mailers on pre-approved loan based on creditworthiness. When one applies for credit, often it’s approved, sometimes it isn’t... Have you ever wondered how exactly the decision to approve a loan is done?  Was the bank able to clearly articulate the rationale of their decision to decline? To ensure fairness and transparency, there are compliance rules which mandates Banks to have such a mechanism. The Federal Trade Commission (FTC), the nation’s(US) consumer protection agency, enforces the Equal Credit Opportunity Act (ECOA), which prohibits credit discrimination on the basis of race, color, religion, national origin, sex, marital status,age, or because you get public assistance. In BFS, often underwriter needs to decide the creditworthiness of an applicant. Among other things, a scorecard is typically in place and long been used. Now with data deluge and better processing power many Analytical models are also used.
Read more

The auxService:mapreduce_shuffle does not exist

Error and Probable Fix: 1. Error : org.apache.hadoop.yarn.exceptions.InvalidAuxServiceException: The auxService:mapreduce_shuffle does not exist Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 1 2015-05-02 13:57:39,982 Stage-1 map = 0%,  reduce = 0% 2015-05-02 13:57:49,511 Stage-1 map = 100%,  reduce = 100% Ended Job = job_1430542224940_0001 with errors Error during job, obtaining debugging information... Examining task ID: task_1430542224940_0001_m_000000 (and more) from job job_1430542224940_0001 Task with the most failures(4): ----- Task ID:   task_1430542224940_0001_m_000000 URL:   http://0.0.0.0:8088/taskdetails.jsp?jobid=job_1430542224940_0001&tipid=task_1430542224940_0001_m_000000 ----- Diagnostic Messages for this Task: Container launch failed for container_1430542224940_0001_01_000005 : org.apache.hadoop.yarn.exceptions.InvalidAuxServiceException: The auxService:mapreduce_shuffle does not exist at sun.reflect.NativeCons
Read more

Zeppelin and Anaconda

Set Anaconda As Default Python Interpreter In Zeppelin Click  anonymous  in top right corner. Click  Interpreter . Scroll down to the  python  interpreter. Click  Edit . Locate  zeppelin.python . Set value to  /home/hadoop/anaconda/bin/python Now find the  spark  interpreter. Locate  zeppelin.pyspark.python . Set value to  /home/hadoop/anaconda/bin/python https://dziganto.github.io/zeppelin/spark/zeppelinhub/emr/anaconda/tensorflow/shiro/s3/theano/bootstrap%20script/EMR-From-Scratch/ This works..tested.. conda install -c calex sklearn-pandas P.S:  zeppelin 0.7.3 don't support spark 2.3. spark 2.3 is supported by zeppelin 0.8 which will be released soon zeppelin 0.7.3 doesn't support python 3.6.. now testing with spark 2.1 and python 3.5...so far so good.. Check:::::: zeppelin home/conf should have zeppelin-env.sh. Here you can specify spark home and zeppelin port Starting Apache Zeppelin from the Command Line On all unix like platforms
Read more