The new version of JMP 18 does not include the sample data set "Boston Housing.jmp" that was previously included with the software. This sample data set was useful for illustrating how to visualize multivariate data and build predictive models, but was removed because it contained discriminatory, unethical and unfair content.
reference
JMP® sample data table "Boston Housing.jmp" removed - JMP User Community
In addition, the Python library "scikit-learn" once included sample data for Boston Housing, but for the same reason it is no longer included. The following is the message that is displayed when importing Boston Housing in Python.
--------------------------------------------------------------------------------------------------------------------------
The scikit-learn maintainers therefore strongly discourage the use of this dataset unless the purpose of the code is to study and educate
about ethical issues in data science and machine learning.
In this special case, you can fetch the dataset from the original
source::
import pandas as pd
import numpy as np
data_url = "http://lib.stat.cmu.edu/datasets/boston"
raw_df = pd.read_csv(data_url, sep="\s+", skiprows=22, header=None)
data = np.hstack([raw_df.values[::2, :], raw_df.values[1::2, :2]])
target = raw_df.values[1::2, 2]
Alternative datasets include the California housing dataset and theAmes housing dataset.
--------------------------------------------------------------------------------------------------------------------------
This message explains that the "California housing dataset" is one of the alternative data for Boston Housing. Not only is it different in that it uses data from the West Coast of the United States instead of the East Coast, but the number of data and the nature of the data are also significantly different.
JMP does not include this California data set as a sample, but you can use the new Python features in JMP 18 to extract this data.Let's load it in JMP.
Python Integration in JMP 18
JMP has long had the ability to connect to Python installed on the same machine, allowing you to execute Python commands and libraries from JMP.
The Python integration feature has been significantly changed and enhanced in the new version, JMP 18. The main changes and enhancements are as follows:
- JMP 18 installs an embedded version of Python (version 3.11) so you do not need to install Python separately from JMP; it will work with the installed Python instance.
- It now has a native Python editor (the JMP Python Script Editor) where you can write Python commands directly (previously, the only way to connect was through Python calls from the JMP Scripting Language (JSL)).
- You can use Python packages (jmp, jmputils) that are useful for using JMP. The jmp package is useful for manipulating data tables (opening, closing, adding/deleting/changing columns, etc.), and the jmputils package is useful for installing packages in the Python environment of JMP.
Opening and Running the JMP Python Script Editor
Launch Python Script from the JMP 18 menu bar and write and run Python programs (Edit > Run Script).
Below is the code to load the scikit-learn dataset in pandas dataframe format as a JMP data table. Paste this code into a JMP Python script andI will do it.
#Install pandas and scikit-learn packages if you have not installed
#import jmputils
#jmputils.jpip('install', 'pandas scikit-learn')
#California dataset
from sklearn.datasets import fetch_california_housing
import pandas as pd
import jmp
california = fetch_california_housing()
# Transform to pandas DataFrame
df_features = pd.DataFrame(data=california.data, columns=california.feature_names)
df_features['MedHouseVal'] = california.target
column_names = df_features.columns.tolist()
# Create JMP Datatable from pandas DataFrame
dt = jmp.DataTable('California',df_features.shape[0])
for i in range( df_features.shape[1] ):
dt.new_column(column_names[i])
dt[i] = list(df_features.iloc[:,i])
# About california_housing_dataset
print(california.DESCR)
Note : Even with JMP's Python, you must install the packages you use in advance. In this example, we use the pandas and scikit-learn packages, so if you do not have them installed, comment out (delete the #) the second and third lines above before running the program.
California Dataset
This is a dataset loaded in JMP. It shows the median home prices for 20,640 blocks in California, which is the objective variable (Y). The explanatory variables include eight items: median income per block, median age of the building, average number of rooms and bedrooms per household, population, number of people per household (average), latitude, and longitude.
Using the latitude and longitude information, I used Graph Builder to plot the home prices (MedHouseVal) for each block on a map. 
Home prices seem to be higher along coastal blocks, especially around larger cities like San Francisco and Los Angeles.
In the next post, Part 2, we will go into more detail about the new features in JMP 18 using the California data set introduced in this post.
by Naohiro Masukawa (JMP Japan)
Naohiro Masukawa - JMP User Community
