Icon

Portillo_​A2_​workflow

Branch 3

Branch 2

Branch 1

Branch 4

CSV Reader
used to find outliers in num medications
Histogram
Used to find the amount of ? a column would have
Value Counter
Used to get each storage type
Extract Table Spec
found the distinct values
GroupBy
better way to look at the distinct values
Table Transposer
Keeps 1 visit per patient
Duplicate Row Filter
flagged extreme valuesin num medications
Numeric Outliers
mean, std deviationwith outlier
Statistics
CSV Writer
Q1 and Q3 without outliers
Box Plot
chi script
Python Script
mean , std dev without outliers
Statistics
Q1 and Q3 with outliers
Box Plot
Remove columns with no change in value
Constant Value Column Filter
info script
Python Script
checks for correlation among columns
Linear Correlation
flag columns with no variation
Low Variance Filter
race
Crosstab
Selects the 8 numeric columns
Column Filter
Apply min-max scaling
Normalizer
labels ? as missing
Rule Engine
Applies z-score
Normalizer
gender
Crosstab
PCA with z-score scaling
PCA Compute
admission type
Crosstab
Value Counter
PCA with no scaling
PCA Compute
gets max num of encounters a patient had
GroupBy
Gets rid of the original numeric columns and keeps the two components
Column Filter
keeps onlyfirst visit perpatient
Duplicate Row Filter
removes patients in hospice or those who passed away
Row Filter
Squeeze 8 numeric columns into PCA 0 and PCA1
PCA
checked row count
Value Counter
combines data from branch 3 andtwo pca components
Column Appender

Nodes

Extensions

Links