[PYTHON] Label each point on the seaborn scatter plot

The 42-year-old uncle who posted for the first time. I changed to seaborn.

trouble

I wrote a scatter plot and didn't know the correspondence between id and each point, so it was not good. When I looked it up, it seems that there is a way to label each point, so I will leave a memorandum.

Implementation method

 -*- coding utf-8 -*-
import pandas as pd
import numpy as np
import seaborn as sns
import matplotlib.pyplot as plt


def label_point(x,y,id,ax):
  df_tmp=pd.concat({'x':x,'y':y,'id':id}, axis='columns')
  for i, point in df_tmp.iterrows():
    ax.text(point['x'], point['y'], point['id'],\
    fontsize=5\
    )


if __name__=='__main__':
  rfilename='testdata.csv'
  data=pd.read_csv(rfilename)

 # Scatter plot drawing
  ax=sns.scatterplot(x=data.iloc[:,3], y=data.iloc[:,4], hue=data.iloc[:,5])

 # Labeling
  label_point(data.iloc[:,3],data.iloc[:,4],data.iloc[:,0],ax)

 #Save
  wfilename='testdraw.png'
  plt.savefig(wfilename)

Part of the data

This is a two-dimensional version of the language characteristics of travel composition, but details are omitted.

id sex age 1-dim 2-dim lab 1 1 63 147.05500793457 -81.8567810058594 me 2 2 45 -128.938018798828 88.1118698120117 fy 3 1 66 20.0744075775146 113.524360656738 me 4 1 68 -64.7453689575195 49.7739143371582 me 5 2 49 -26.7232112884521 -164.791641235352 fy

Drawing result

The id is a floating point number, but it's not the main subject, so leave it as it is.