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

# # <img src="images/data_logo.png"></img>
# 
# A few examples of using Lets-Plot with dictionaries, Pandas DataFrames and Polars DataFrames.

# ### Table of Contents
# 
# 1. [Python Dictionaries](#dict)
# 
# 2. [Pandas Dataframe](#pandas)
# 
#     2.1. [From Dictionary](#pandas-from-dict)
# 
#     2.2. [From CSV](#pandas-from-csv)
# 
# 3. [Polars Dataframe](#polars)
# 
#     3.1. [From Dictionary](#polars-from-dict)
# 
#     3.2. [From CSV](#polars-from-csv)

# In[1]:


import numpy as np
import pandas as pd
import polars as pl

from lets_plot import *


# In[2]:


LetsPlot.setup_html()


# <a id="dict"></a>
# 
# ### 1. Python Dictionaries

# In[3]:


x = np.linspace(-2 * np.pi, 2 * np.pi, 100)
y = np.sin(x)


# In[4]:


ggplot({'x': x, 'y': y}, aes('x', 'y')) + geom_point()


# <a id="pandas"></a>
# 
# ### 2. Pandas Dataframe

# <a id="pandas-from-dict"></a>
# 
# #### 2.1. From Dictionary

# In[5]:


def get_data_dict():
    np.random.seed(42)
    n = 100
    x = np.random.uniform(-1, 1, size=n)
    y = 25 * x ** 2 + np.random.normal(size=n)
    return {'x': x, 'y': y}


# In[6]:


pandas_df = pd.DataFrame(get_data_dict())
print(pandas_df.shape)
pandas_df.head()


# In[7]:


ggplot(pandas_df) + \
    geom_point(aes('x', 'y', fill='y'), shape=21, size=5, color='white')


# <a id="pandas-from-csv"></a>
# 
# #### 2.2. From CSV

# In[8]:


# Load mpg dataset with pandas

mpg_pandas_df = pd.read_csv("https://raw.githubusercontent.com/JetBrains/lets-plot-docs/master/data/mpg.csv")


# In[9]:


ggplot(mpg_pandas_df, aes('displ', 'cty', fill='drv', size='hwy')) + \
    geom_point(shape=21) + \
    scale_size(range=[5, 15], breaks=[15, 40]) + \
    ggsize(600, 350)


# <a id="polars"></a>
# 
# ### 3. Polars Dataframe

# <a id="polars-from-dict"></a>
# 
# #### 3.1. From Dictionary

# In[10]:


polars_df = pl.DataFrame(get_data_dict())
print(polars_df.shape)
polars_df.head()


# In[11]:


ggplot(polars_df) + \
    geom_point(aes('x', 'y', fill='y'), shape=21, size=5, color='white')


# <a id="polars-from-csv"></a>
# 
# #### 3.2. From CSV

# In[12]:


# Load mpg dataset with polars

mpg_polars_df = pl.read_csv("https://raw.githubusercontent.com/JetBrains/lets-plot-docs/master/data/mpg.csv")


# In[13]:


ggplot(mpg_polars_df, aes('displ', 'cty', fill='drv', size='hwy')) + \
    geom_point(shape=21) + \
    scale_size(range=[5, 15], breaks=[15, 40]) + \
    ggsize(600, 350)