Introduction In this post I want to take a look at my Netflix viewing habits. The first step is getting the data and you can request your viewing data from the Accounts section in your Netflix account. Netflix will allow you to download a zip file with many different ways to slice this information. # Import the libraries import pandas as pd import numpy as np import seaborn as sns from pandas.

Continue reading

Introduction In this post, we are going to look at some publicly available data to dig deeper into exploratory data analysis and machine learning techniques. We’ll look at some data from Jonathan McDowell’s Catalog site. This site has data about man-made objects launched into space and all the debris we have created in near earth orbits. We’ll try to load the file directly from the site. This allows us to get the most recent data from the internet.

Continue reading

library(tidyverse) library(ambient) library(randomcoloR) library(ggforce) disturbance = expand.grid(c = 1:25, r = 1:49) %>% mutate( c = ifelse(r %% 2 == 0, c + 0.5, c), a = 180 * gen_cubic(c, r, frequency = 0.1, seed = 1964) ) %>% filter(c <= 50) ggplot(disturbance) + geom_text(aes(c, r, label = "0", angle = a, color = a), family = "Times", size = 16, show.legend = F) + coord_fixed(ratio = 0.5, expand = TRUE) + scale_color_gradient2(low = "orange", high = "red", mid = "tomato") + theme_void() set.

Continue reading

Introduction In this post, we are going to look at some publicly available data to dig deeper into exploratory data analysis and machine learning techniques. I am going to start by fetching some data from the inter webs, this data is available at the FuelEconomy.gov site. This file has fuel economy data for all cars sold in the United States for several years. Let’s start by loading the libraries we need:

Continue reading

NPS analysis NPS - Comment analysis In an [previous post](https://nitinahuja.github.io/2017/nps-exploratory-analysis-in-r/) we performed some EDA on the NPS data we have. Recall that as part of the question about the likelihood of recommending a service or business there is an optional text response about why they picked this score. Let’s try and see what those responses are all about. We had already performed some sentiment analysis on this text we are now going to attempt to classify this text into topics.

Continue reading

Heatmaps Heat maps are invaluable in displaying a large amount of continuous data contained in a 2d matrix. This post is meant to show a way to create a print worthy heat map in R. Let’s start by loading the required packages. suppressPackageStartupMessages({ library(ggplot2) library(ggthemes) library(viridis) library(scales) library(tidyr) }) Data Our data is from a business that receives sales calls 24x7. Let’s read and see what the data looks like. We have observations (count of calls) for each day of the week and each hour of the day.

Continue reading

When you are managing managers, you should have a single focus; they are learning to manage their teams well and as a secondary objective that they are contributing at high levels. There are several experts in this area and I do not claim to be anywhere close; these are just reminders for myself. Your team learns by modeling you. There is little doubt that what you do, your team will emulate.

Continue reading