Data Analysis of an International Hardware Sales Company Using SQL

Instagram is one of the most popular social media platforms with over 1 billion monthly active users which means it’s a great marketing channel to use for promoting your brand. That's why I tried to make a dashboard with which I can analyze the performance of the Instagram accounts That i want

[Project Github Repository](https://github.com/meysamraz/instagram-analytics-dashboard-streamlit)

Watch Demo here :
https://instagramanalytic.herokuapp.com/

### Who can use this dashboard  ?

*   Page  Admin  :   This dashboard can help to measure their performance and optimize their page on instagram
*   Markting teams : Using this dashboard they can see what's popular and get a better understand of their audience
*   Investors : Presented metrics on this dashboard can give a better understanding of our performance on Instagram
*   Manager : Using this dashboard, the manager can monitor the performance of the Instagram team or admin

### Whats our metrices on this dashboard ?

*   Total growth of followers per day/weak/month/year
*   Total growth of followings per day/weak/month/year
*   Total growth of posts per day/weak/month/year
*   Total growth of likes per day/weak/month/year
*   Total growth of comments per day/weak/month/year
*   Total growth of videos views per day/weak/month/year

**You also can comapre and analyze posts to each other and see which posts perform better**

### How this Dashboard works :

#### 1 - Provide the data :

I used a python package called [instaloader](https://instaloader.github.io/) to collect data every 24h per day

### 2 - Front Side and Charts :

I used a powerful framework [Streamlit](https://streamlit.io/) to build (front side all in python) and  [plotly](https://plotly.com/) and [matplotlib](https://matplotlib.org/) to build the charts

### 3 - Server and Database :

I used Heroku a cloud platform as a service which provide a free hosting to deploy my app on it. it's and amzaing platform gave me so much flexbilte to deploy my apps and provide me an free PostgreSQL database and a scheduled service which i used to run my data collecting scripts every 24h all data will be collected 24h and insert into my Heroku PostgreSQL database

**The Heroku database service has a storage space limit, so I had to stop my data collection operation due to the lack of storage space.**

## How to run the project :

Install requirements :

`pip install -r requirements.txt`

Run the project :

`streamlit run main.py`

## Libraries and FrameWorks used in the project

*   [streamlit](https://streamlit.io/)
*   [pandas](https://pandas.pydata.org/)
*   [numpy](https://numpy.org/)
*   [plotly](https://plotly.com/)
*   [matplotlib](https://matplotlib.org/)
*   [psycopg2](https://pypi.org/project/psycopg2/)
*   [SQLAlchemy](https://sqlalchemy.org/)



Instagram Dynamic Analytics Dashboard Streamlit and Heroku

## Introduction

![](https://cdn1.dotesports.com/wp-content/uploads/2022/05/10133527/you.tube\_.logo\_.4.jpg)

YouTube is a very rich educational resource for data science and related fields. In this project, I tried to extract the data of four of my favorite channels from Yotube API. These data are related to comments, channel information, and video information, and then From extracting their performance, I analyze and compare them

Tip - The visualizations and analyzes used in this project can be used to analyze any channel, just extract the desired channel data.

### Open notebook from my [github repository](https://github.com/meysamraz/youtube-api-exploratory-data-analysis)

In this project, I will cover and explore the following:

* ### Extract Data from Youtube API .

* ### Data Preprocessing and Feature Engineering

*   ### Analyzing (channel,videos,comments) data and analyze and compare their performance and see what made them popular, for example :
    *   How many subscribers they have and how far are they from each other ?

    *   Which one has uploaded more videos ?

    *   Comparison of the number of views of each channel

    *   How long did it take to reach their current position ?

    *   The most viewed video of each channel

    *   How long are the videos usually?

    *   What day do they usually upload videos ?

    *   In which year did Chanel perform better ?

    *   Whats average duration of videos

    *   check correlation :

    *   Dose video durition matter for views ?

    *   Does title length matter for views ?

    *   Does likesCount and commentsCount correlated with views ?
*   ### Explore text data using NLP techniques
    *   Most written in the comments for each channel
    *   Most words used in titles
    *   Sentiment analysis for the comments of each channel

## Data Visualization And Exploratory analysis preview

![](/images/sentiment_analysis.png)![](/images/views.png)![](/images/corr.png)


Exploratory Data Analysing My Favorite Data Science Channels Using Youtube API

[Project Github repository
](https://github.com/meysamraz/digikala-customer-purchase-dashboard)

Digikala is an popular Iranian e-commerce company I personally buy a lot from them and in this project I tried to create a dashboard to analyze my purchases.

## Project Overview :



### 1 - Collecting Data :

I used Digikala's hidden api to collect data, fortunately Digikala uses api to load data and I didn't have much trouble collecting data (There was no scraping :) )

#### What data did I collect ?

*   Orders data

    *   id

    *   payable_price

    *   created_at

    *   total_cost

    *   shipping_cost

    *   discount

*   purchased product data

    *   id

    *   order_id

    *   created_at

    *   title_fa

    *   title_en

    *   image_url

    *   category

### 2 - Preprocess and Feature engineering collected data :

Convert order date from Jalali to Gregorian

*   Clean category text

*   save datas into csv

### 3 - Creating the dashboard using Power BI

For creating dashboard i used Power BI an powerful tool and interactive data visualization software product developed by Microsoft

#### What i persented in my dashboard :

**Note: that all the monetary units presented in this dashboard are in Rials**

*   Total Expense (Collection of money spent in Digikala)

*   Purchased Products (Number of products purchased)

*   Orders Count (Number of times ordered)

*   Return Products (Number of products purchased and returned)

*   Total Spent by date (Total spent by year and month)

*   Top 5 purchased categories (Which categories were the most purchased?)

*   5 most expensive products purchased (The most expensive products bought)

### Libraries and software used :

*   [Power BI](https://powerbi.microsoft.com/en/)

*   [requests](https://pypi.org/project/requests/)

*   [pandas](https://pandas.pydata.org/)

*   [persiantools](https://pypi.org/project/persiantools/)

*   [deep-translator](https://pypi.org/project/deep-translator/)



Collecting and Dashboarding Customer Purchase Data from a Famous E-commerce Website

layout

By default, the <title> tag for this page is determined by the title field (in the Content group). You can override the tag value here.

metaTitle

media

bottomSections

By default, the description tag for this post is taken from the Excerpt field. You can override the tag value here.

metaDescription

By default, the "og:image" meta tag set for social sharing this post points to the Featured Image field (in the Content group). You can override that image here.

socialImage

markdown_content

If enabled, the title suffix defined in the site configuration is appended to the title tag of this page.

addTitleSuffix

To add or override any meta tag for this page, add entries to this list. Entries defined here take precedence over any other defaults.

metaTags

title

colors

date

author

excerpt

featuredImage

##### [Project Github repository](https://github.com/meysamraz/analyzing-amazon-books-related-to-data-science)

In this project, I analyzed the data collected from tweets related to the World Cup hashtag using the Twitter API. The purpose of this analysis is to get to know the tweet pattern and find the trends of the Qatar World Cup.

## Project Overview

![](/images/plot3.png)

![](/images/plot5-4ec62bd9.png)

*   ### Collect data from Twitter
*   ### Understanding the data
    *   shape of the data
    *   check column dtypes
    *   check is there any null values
*   ### Extract user mentioneds
*   ### Top 20 most mentiond users
*   ### Most tweets belongs to who ?
*   ### Min and Max and Mean Retweet Count
*   ### Min and Max and Mean Followers Count and Friends Count
*   ### Which device are the most tweets done with ?
*   ### Find accounts that created when world cup started
    *   How many accounts were created during the World Cup ?
    *   What's the language of recent accounts ?
    *   What places did they tweet from the most ?
    *   Most used source for tweeting ?
    *   What was the maximum number of followers they could attract during the World Cup?

### Libraries used in the project

*   [pandas](https://pandas.pydata.org/)
*   [numpy](https://numpy.org/)
*   [matplotlib](https://matplotlib.org/)
*   [seaborn](https://seaborn.pydata.org/)
*   [termcolor](https://pypi.org/project/termcolor/)
*   [wordcloud](https://pypi.org/project/wordcloud/)
*   [tweepy](https://www.tweepy.org/)



Collecting and EEDA of World Cup Qatar Tweets

##### [Github repository](https://github.com/meysamraz/RFM-customer-segmentation-retail-shop)

##### [Project Analysis Report](https://docs.google.com/document/d/1usBhCDgDJz3gsrSh9Ve0VTt8pMpNhxd3zg-5kaotyBk/edit?usp=sharing)

An e-commerce company wants to segment its customers and determine marketing strategies according to these segments. To this end, we will define customers' behavior and create groups according to clusters in these behaviors. In other words, we will include those who exhibit common behaviors in the same groups and we will try to develop special sales and marketing techniques for these groups.

I'm going to use RFM which stands for Recency, Frequency, and Monetary Value. It is a marketing framework that is used to analyze customer behavior and segment customers into groups based on their past purchasing habits.

The RFM analysis is based on the following three metrics:

Recency - How recently a customer has made a purchase.
Frequency - How often a customer makes purchases.
Monetary Value - How much a customer spends on each purchase.

The RFM formula involves assigning a score to each metric, which is then used to segment customers into different groups. The scores are typically ranked on a scale of 1 to 5, with 5 being the highest score.

The RFM formula can be written as follows:

RFM Score = (R Score * 100) + (F Score * 10) + M Score

Where:
R Score = a score based on how recently the customer made a purchase (e.g., 5 for most recent, 1 for least recent)
F Score = a score based on how frequently the customer makes purchases (e.g., 5 for most frequent, 1 for least frequent)
M Score = a score based on how much the customer spends on each purchase (e.g., 5 for highest spend, 1 for lowest spend)

By segmenting customers into different groups based on their RFM scores, businesses can tailor their marketing strategies and offers to better meet the needs and preferences of each group, ultimately leading to increased customer loyalty and revenue.

![](/images/plt2.png)![](/images/plt0.png)![](/images/plt5.png)![](/images/plt1.png)

![](/images/plt4.png)



*   Dataset
    *   InvoiceNo: Invoice number. Nominal. A 6-digit integral number is uniquely assigned to each transaction. If this code starts with the letter 'c', it indicates a ---cancellation.
    *   StockCode: Product (item) code. Nominal. A 5-digit integral number is uniquely assigned to each specific product.
    *   Description: Product (item) name. Nominal.
    *   Quantity: The quantities of each product (item) per transaction. Numeric.
    *   InvoiceDate: Invoice date and time. Numeric. The day and time when a transaction was generated.
    *   UnitPrice: Unit price. Numeric. Product price per unit in sterling ( £).
    *   CustomerID: Customer number. Nominal. A 5-digit integral number is uniquely assigned to each customer.
    *   Country: Country name. Nominal. The name of the country where a customer resides.

### Project Overview

#### 1 -  Exploring Dataset

*   Check data shape
*   Check data type
*   See if there is any null values
*   See if customers have frequency in their purchases
*   Check the first and last recorded purchases

#### 2 - Data Preprocessing

*   Drop null values
*   Only keep recoreds with real purchasing
*   Create a new column call ***TotalAmount*** shows total amount of money customer spent on order
*   Reformat InvoiceDate to datetime

#### 3 - Calculating RFM Score

Calculating Recency

*   Calculating Frequency
*   Calculating Monetary
*   Score R and F and M based on their quantile
*   Segment customers based on their RFM score max = 555 and min = 111

#### 4 - Analyzing Segments

*   How many customers belong to each Segment?
*   AVG customers total amount spent
*   What's Yearly total amount and quantity sold?
*   What country are the customers from?
*   What country are the champion customers from?
*   Most Frequently bought product by customers of each segment

#### 5 - Reporting Analysis

In this report, I summarize the analysis that has been made in this notebook + Explaining each segment and marketing strategy for them.

*   #### https://docs.google.com/document/d/1usBhCDgDJz3gsrSh9Ve0VTt8pMpNhxd3zg-5kaotyBk/edit?usp=sharing

### Libraries used in the project

*   [pandas](https://pandas.pydata.org/)
*   [plotly](https://plotly.com/)
*   [matplotlib](https://matplotlib.org/)
*   [seaborn](https://seaborn.pydata.org/)



RFM Customer Segmentation For a Retail Shop

##### [Project Github repository](https://github.com/meysamraz/electric-cars-tracker-dashboard-US)

Electric cars become so popular this days especially in the U.S . In this project I made a dashboard to track electric cars across the U.S using powerBI . This dashboard is full interactive.

The data that I used for dashboard is available on Kaggle :
[Link of the dataset](https://www.kaggle.com/datasets/ratikkakkar/electric-vehicle-population-data)

*   VIN (1-10)
    Vehicle Identification Number of the vehicle mentioned in the dataset.

*   County :
    Name of the County from where the data is gathered.

*   City :
    Name of the Cities from where the data is gathered.

*   State :
    Name of the State from where the data is gathered.

*   Postal Code :
    The postal code from where the data is gathered.

*   Model Year :
    Manufacturing year of the model mentioned in the data set.

*   Make :
    Manufacturer of the vehicle.

*   Model :
    Model Name of the mentioned vehicle.

*   Electric Vehicle Type :
    Type of the vehicle present in the dataset.

## Features of This Dashboard :

By using this dashboard you be able to track

*   Number of active electirc cars across the U.S

*   The most purchased car brand

*   State with the most active cars

*   City of state which has the most electric cars

*   States where electric cars are available

*   Type of Electric types

*   Year of manufacture of existing electric cars

#### You are also able to filter data by **state**,**Brand**,and **Type**

## Libraries and software used :

*   [Power BI](https://powerbi.microsoft.com/en/)


Interactive Electric Cars Tracker Dashboard The US version

##### [Project Github repository](https://github.com/meysamraz/ab-testing-mobile-game)



In this project i will walk through an ab testing for an popular mobile game app i will show step-by-step of an ab testing using python

Etiam facilisis lacus nec pretium lobortis. Praesent dapibus justo non efficitur efficitur. Nullam viverra justo arcu, eget egestas tortor pretium id. Sed imperdiet mattis eleifend. Vivamus suscipit et neque imperdiet venenatis. In malesuada sed urna eget vehicula. Donec fermentum tortor sit amet nisl elementum fringilla. Pellentesque dapibus suscipit faucibus. Nullam malesuada sed urna quis rutrum. Donec facilisis lorem id maximus mattis. Vestibulum quis elit magna. Vestibulum accumsan blandit consequat. Phasellus quis posuere quam.

This project is based on a mini project from <a href="https://www.datacamp.com/projects/184">Datacamp</a>.
<a href="https://www.facebook.com/cookiecatsgame">Cookie Cats</a> is a hugely popular mobile puzzle game developed by <a href="http://tactile.dk">Tactile Entertainment</a>. It's a classic "connect three"-style puzzle game where the player must connect tiles of the same color to clear the board and win the level. It also features singing cats. Check out this short demo:
<a href="https://youtu.be/GaP5f0jVTWE"><img src="https://s3.amazonaws.com/assets.datacamp.com/production/project_184/img/cookie_cats_video.jpeg" style="width: 500px"></a>
As players progress through the levels of the game, they will occasionally encounter gates that force them to wait a non-trivial amount of time or make an in-app purchase to progress. In addition to driving in-app purchases, these gates serve the important purpose of giving players an enforced break from playing the game, hopefully resulting in that the player's enjoyment of the game being increased and prolonged.But where should the gates be placed? Initially the first gate was placed at level 30. In this project, we're going to analyze an AB-test where we moved the first gate in Cookie Cats from level 30 to level 40. In particular, we will look at the impact on player retention. 
<img src="https://s3.amazonaws.com/assets.datacamp.com/production/project_184/img/cc_gates.png" alt="">

The data is from 90,189 players that installed the game while the AB-test was running. The variables are:
<ul>
<li><code>userid</code> - a unique number that identifies each player.</li>
<li><code>version</code> - whether the player was put in the control group (<code>gate_30</code> - a gate at level 30) or the test group (<code>gate_40</code> - a gate at level 40).</li>
<li><code>sum_gamerounds</code> - the number of game rounds played by the player during the first week after installation
<li><code>retention_1</code> - did the player come back and play 1 day after installing?</li>
<li><code>retention_7</code> - did the player come back and play 7 days after installing?</li>
</ul>
When a player installed the game, he or she was randomly assigned to either <code>gate_30</code> or <code>gate_40</code>. 

### AB Testing Process

1. Understanding business problem & data
2. Detect and resolve problems in the data (Missing Value, Outliers, Unexpected Value)
3. Look summary stats and plots
4. Apply hypothesis testing and check assumptions
 * Check Normality & Homogeneity
 * Apply tests (Shapiro, Levene Test, T-Test, Welch Test, Mann Whitney U Test)
5. Evaluate the results
6. Make inferences
7. Recommend business decision to your customer/director/ceo etc.

# Libraries and packages :

* [google-api-python-client](https://pypi.org/project/google-api-python-client/)
* [pandas](https://pandas.pydata.org/)
* [numpy](https://numpy.org/)
* [nltk](https://www.nltk.org/)
* [matplotlib](https://matplotlib.org/)
* [seaborn](https://seaborn.pydata.org/)
* [wordcloud](https://amueller.github.io/word_cloud/)
* [termcolor](https://pypi.org/project/termcolor/)

AB Testing Mobile Game Feature

[Project Github repository](https://github.com/meysamraz/linkedin-profile-dashboard-power-bi)

# Overvirew

LinkedIn is a very powerful social network for communicating and finding a job. In this project, I tried to analyze my profile using the data provided by LinkedIn. This project is not only for analyzing my profile, and anyone who is active in LinkedIn can Use this dashboard.

I used [Power BI](https://powerbi.microsoft.com/en/) a great data visualisation tool published by microsoft Power BI provied us great Data trasformation tool to transform our data and easily visualize it.

But before i start, I used at the end i used a little python code to understand better data (optional) :)

## Where i get my data :

Fortunately, LinkedIn has a suitable platform for obtaining information about ourselves. That means there is no scraping :)

Open your LinkedIn -> profile -> setting -> data privacy -> Get a copy of your data

![](/images/linkedin_data_pull.png)

### Libraries and packages for python notebook :

*   [pandas](https://pandas.pydata.org/)



Linkedin Profile Dashboard Power BI

In this project, I tried to detect the traffic on both sides of the road using YOLO algorithm and opencv. All detected data are stored in mongodb database for analysis. The user interface environment of this application is made using streamlit

##### [Project Github repository](https://github.com/meysamraz/alzheimers_diagnosis_using_brain_MRI)

![](/images/demo.gif)

![](/images/mongo_env.png)

## Libraries and FrameWorks used in the project

*   [streamlit](https://streamlit.io/)

*   [numpy](https://numpy.org/)

*   [matplotlib](https://matplotlib.org/)

*   [opencv](https://opencv.org/)

*   [pymongo](https://pymongo.readthedocs.io/)

*   [pytorch](https://pytorch.org/)



My Projects