Skip to main content

Top 10 basic command you should try in your data analysis / Machine learning




Hi all,

When you get the data in CSV format or any other format then you should try this 10 basic command to understand more the data. How much you understand the dataset will help you to make better model for it.

Before jumping into the code I want to mention some basic packages/software you want to have:

  1. Python 
  2. Jupyter Notebook (Highly recommend this one) or Spyder
  3. pandas (default it will be installed in python)
That's it. These are the basic things you should need to start. First get some the data from kaggle or anywhere.Let the data format will be CSV format for better understanding.

Import the necessary packages and data like this

import pandas as pd
df= pd.read_csv("path to csv file)


Now Lets the establish top common commands:

1. df.head(): This command will show top rows in the dataset

2. df.info(): This command will tell the type of each column such as float, int, object type.

3. df. describe(): This will describe the column in the form as min, max, standard deviation, mean,          count. This command is used to under the relation between each column in numerical form.

4. df.shape: This is used to know the number of column and no.of rows in the data.

5. df.value_counts(): This command I use most of the time. This command is used to the count of            each unique value in the column.

6. df.unique(): This used to know the unique value in the column.

7. df.groupby: This somewhat big command compare to the others. This is used to group two                  columns or group by some condition to make another data frame.

8. df.sort_values(): This used to sort the value of the column-like ascending or descending order.

9. df.isnul().sum(): This is an important command. This command will help us know the number of        null values in each column

10. df.fillna(): This command you may predict. This is used to fill the null values with any                        variable  or   string.

I fell these are the some repeated pandas command we use in our day to day work to know the data more about it. If you have any other commands or suggestion feel free to post it in the comments section guys.

Happy Learning!!!!









Comments

Post a Comment

Popular posts from this blog

Welcome

Hi everyone, This is Karthick Aravindan, I am a  Machine Learning Engineer working in Optisol Business Solution. I just want to share my experience in this field. This blog is completely dedicated to machine learning stuff. As into to blog, I am sharing my daily tools what I am using for my work. 1) Python 3.+ version:  https://www.python.org/ 2) Anaconda or miniconda both are best install anyone:  https://www.anaconda.com/download/ 3) Jupyter notebook 4) Colabs form google, I will write a sperate post for this: https://colab.research.google.com 5) Pycharm awesome IDE for python:  https://www.jetbrains.com/pycharm/download/ 6) AWS account to train module or use colab to train quickly. 7) Any text editor like sublime, atom, vscode. I recommend vscode it is really cool stuff to have on         your desk:  https://code.visualstudio.com/ I feel these are the basic things we need to...
Post a Comment
Read more

How to load a dataset in Python?

This is the first step for Data scientist work. Data is very important in this field without data there is nothing to do here. I mean data is everything, not only number it consists of image, audio,  video or anything. In this post, I will tell the process to load simple dataset in CSV format. I recommend installing jupyter or spyder from Anaconda website. Packages you want to install in cmd or terminal  pip install pandas pip install numpy pip install matplotlib Mostly these packages will be installed default. Just for a verification check whether  it is installed. Numpy             It is a mathematical tool used to do some complex calculation. Pandas              Without pandas, we can import the data. It is manually used to form a data frame and other allocation processes. Matplotlib                 It ...
Post a Comment
Read more