#!/usr/bin/env python
# coding: utf-8

# In[1]:


import numpy as np
import random
from lets_plot import *

LetsPlot.setup_html()


# In[2]:


# This example was found at: www.cookbook-r.com/Graphs/Scatterplots_(ggplot2)
random.seed(123)
data = dict(
    cond=np.repeat(['A','B'], 10),
    xvar=[i + random.normalvariate(0, 3) for i in range(0,20)],
    yvar=[i + random.normalvariate(0, 3) for i in range(0,20)]
)


# #### Basic scatter-plot

# In[3]:


p = ggplot(data, aes(x='xvar', y='yvar')) + ggsize(300, 250)
p + geom_point(shape=1) 


# #### Add regression line

# In[4]:


p + geom_point(shape=1) + geom_smooth() 


# In[5]:


# Without standard error band.
p + geom_point(shape=1) + geom_smooth(se=False) 


# #### Split dataset by the `cond` variable

# In[6]:


p1 = ggplot(data, aes(x='xvar', y='yvar', color='cond')) + ggsize(500, 250)
p1 + geom_point(shape=1) + geom_smooth(se=False)


# In[7]:


# Map `shape` to the `cond` variable.
p2 = p1 + geom_point(aes(shape='cond'), size=5)
p2


# In[8]:


# Choose different shapes using `scale_shape_manual`:
# 1 - hollow circle 
# 2 - hollow triangle

p2 + scale_shape_manual(values=[1,2])


# #### Handling overplotting

# In[9]:


# Create data containing overlapping points.
data['xrnd'] = [round(v / 5) * 5 for v in data['xvar']]
data['yrnd'] = [round(v / 5) * 5 for v in data['yvar']]


# In[10]:


p3 = ggplot(data, aes(x='xrnd', y='yrnd')) + ggsize(500, 250)

# Use `alpha` to show overplotting.
p3 + geom_point(alpha=.3, size=5)


# In[11]:


# `jitter` points to show overplotting in another way.
p3 + geom_point(shape=1, position=position_jitter(width=.2,height=.5))