How to use ReNom TDA GUI

An introduction of how to use ReNom TDA GUI.

In this tutorial, we explain how to use ReNom TDA GUI. you can learn following points.

  • How to start ReNom TDA GUI.
  • How to use ReNom TDA GUI.

How to start ReNom TDA GUI

To start ReNom TDA GUI, we move to ReNom directory.

And start ReNom TDA GUI server.

cd path to ReNom directory/
python gui/server.py

Open http://0.0.0.0:8080 in browser.(Except IE)

We can show next page.
In this page, we can show statistic values and histogram of data.
gui1

gui1

Load csv file

Next we load csv file.

You can load csv files in ReNom/gui/data directory.

In this directory, there are three sample datastes, finance.csv, hitter.csv and digit.csv.

In this tutorial, load digit.csv.

gui2

gui2

Select calculate columns and colorize columns

Next we select calculate columns and colorize columns.

In this case, columns except target are used to calculate, and all columns are used to colorize.

And go next page.

gui3

gui3

Visualizing page

Next we can show follow page.

This page is visualizing page.

gui4

gui4

Set parameter

Next, we set parameter to visualize.

Push Column 0 tab, we can show three selectbox, dimension reduction algorithm, analysis mode and colored by.

gui5

gui5

gui6

gui6

gui7

gui7

Scatter plot

For example, select PCA, Scatter Plot and target.

we can show follow scatter plot.

When analysis mode is Scatter Plot, we can visualize point cloud data.

gui8

gui8

Unsupervised clustering

Next, select PCA, Clustering(教師なし) and clustering algorithm is K-means.

We can show following.

In this case, we can show the result of unsupervised clustering.

gui9

gui9

Supervised clustering

Next, select PCA, Clustering(教師あり) and clustering algorithm is K-NearestNeighbor.

We can show following.

In this case, we can show the result of supervised clustering.

In supervised clustering, data are splited in train or test.

We can select ratio of train data and train data was colored by transparent color.

gui10

gui10

TDA

Next, select PCA, TDA and colored by target.

We can show following.

In this case, we can show the result of TDA.

gui11

gui11

Split window

We can split visualize window.

In this mode, you can compare many parameters.

gui12

gui12

Show histogram and 3D visualize of TDA

We can show histogram of coloring data and we can show tda result of 3D visualize mode.

In this mode, you can draggable tda nodes.

And we can know near nodes isn’t connect each other.

gui13

gui13

Search data

Next we seach data in topology.

We can select text search or number search and select some operator “=”, “>” or “<”.

gui14

gui14

For example, search target = 3 data in topology.

We can show follow result.

gui15

gui15