import pandas as pd
from matplotlib import pyplot as plt
import seaborn as sns

url = "https://raw.githubusercontent.com/PhilChodrow/PIC16B/master/datasets/palmer_penguins.csv"
penguins = pd.read_csv(url)

penguins.head()

penguins.nunique()

penguins["Sex"].unique()

penguins.index[penguins["Sex"] == "."].tolist()

nan_values_summary = penguins.isna().sum()
nan_values_summary

fig, ax = plt.subplots(1)
# create the bar plot
ax = nan_values_summary.plot.bar(figsize=(16,5))
x.set(xlabel = "Column Name",
       ylabel = "Nan Counts",
       title = "Bar Chart of the Missing Data")

ax.legend(labels = ['Nan Values'],
          bbox_to_anchor = (0.12, 1))

# remove border
ax.spines['right'].set_visible(False)
ax.spines['top'].set_visible(False)

# I find how to add annotate on each bar online from 
# https://www.geeksforgeeks.org/how-to-annotate-bars-in-barplot-with-matplotlib-in-python/
# Iterrating over the bars one-by-one
for bar in ax.patches:
   
  # Using Matplotlib's annotate function and
  # passing the coordinates where the annotation shall be done
  # x-coordinate: bar.get_x() + bar.get_width() / 2
  # y-coordinate: bar.get_height()
  # free space to be left to make graph pleasing: (0, 8)
  # ha and va stand for the horizontal and vertical alignment
    ax.annotate(bar.get_height(),
               (bar.get_x() + bar.get_width() / 2 ,
                bar.get_height()), ha='center', va='center',
                size=10,
                xytext=(0, 6),
                textcoords='offset points',
                color="firebrick") 

# Drop columns 
drop_list = ["Region", "Stage", "studyName", "Sample Number", "Individual ID", "Comments","Delta 15 N (o/oo)", "Delta 13 C (o/oo)", "Clutch Completion"]

def clean_penguins_data(data_df, drop_list):
    """
    This function will shorten the name of the Penguin Species, 
    drop unuseful columns, and clean the missing (nan) values.
    
    Parameters:
    data_df: data frame to be cleaned
    drop_list: a list of column names I will drop from the input data frame
    
    Return:
    A cleaned dataframe df.
    """
    
    df = data_df.copy() # avoid polluting original data set
    
    # Shorten the species name
    df["Species"] = df["Species"].str.split().str.get(0)
    
    # Remove the entries where sex was not recorded
    df = df[penguins["Sex"] != "."]

    #Drop columns 
    df = df.drop(drop_list, axis = 1)
    
    #Find and drop the rows with nan values
    nan_df = df.isna()
    nan_columns = nan_df.any()
    columns_with_nan = df.columns[nan_columns].tolist() 
    df = df.dropna(subset = columns_with_nan)

    return df

df = clean_penguins_data(data_df = penguins, drop_list = drop_list)
df

df["Date Egg"] = pd.to_datetime(df["Date Egg"])
df['year'] = df['Date Egg'].dt.year
df = df.sort_values(by=["Date Egg"])
df

# colour map for species
species_mapper = {
    "Adelie" : "red",
    "Gentoo" : "green",
    "Chinstrap":"purple"
}

# create the 5 plots, for different species from different islands.
fig, axes = plt.subplots(1, 5, figsize = (30,10), sharey = True, sharex = True)
axes[0].set_ylabel("Date Egg", fontsize = 15) 
ax_list = axes.tolist()                       # get the list of axes
   
def plot_hist(df, colname, alpha):
    """
    This function is used in the apply() method; 
    it helps draw the histogram of "Date Egg" for each species.
    
    Parameters
    ----------
    df: data frame; 
    colname: string; "Date Egg" column
    alpha: float; a user-specified number for transparency

    Return 
    ----------
    No return value 
    """

    specie_name = df["Species"].unique()[0]   # get species Island name 
    island_name = df["Island"].unique()[0]    # get current Island name 
    ax = ax_list.pop(0)                       # get an axis 
    # set title name for each plot
    ax.set_title(specie_name + " - " + island_name, fontsize = 20)
    ax.hist(df[colname],
            alpha = alpha,
            label = specie_name,
            color = species_mapper[specie_name],
            orientation ='horizontal')
    ax.set_xlabel('Number of Eggs', size=17)
    ax.legend(fontsize = 15)

# groupby method with apply
df.groupby(["Species", "Island"]).apply(plot_hist, "Date Egg", 0.5 )

# add a suptitle to the figure.
fig.suptitle("Histogram of 'Date Egg' ",
             fontweight ="bold", size=26)
# show the plot
plt.show()

fgrid = sns.relplot(data=df,
                    x = "Date Egg",
                    y = "Body Mass (g)",
                    hue = "Species",
                    col = "Island")

fgrid.fig.suptitle("Scatterplot of 'Date Egg' Against 'Body Mass (g)' ", size=16)
fgrid.fig.subplots_adjust(top = 0.85)
# turn 2D into 1d, easier to iterate
axes = fgrid.axes.flatten()
# rotate xticks for each plot
for ax in axes:
    ax.tick_params(axis="x", labelsize=10, labelrotation=40, labelcolor="firebrick")

from plotly import express as px
import plotly.io as pio

# pio.templates.default = "ggplot2"

pio.templates.default = "plotly_white"

fig = px.scatter(data_frame = df,      # data set
                 x = "Date Egg",       # column for x axis
                 y = "Body Mass (g)",  # column for y axis
                 color = "Species",    # column for dot color
                 width = 900,          # width of figure
                 height = 700,         # height of figure
                 opacity = 0.6,        # transparency for dot
                 facet_col = "Sex",    # assign marks to facetted subplots in the horizontal direction.
                 facet_row = "Island", # assign marks to facetted subplots in the vertical direction.
                 title = "Scatterplot of 'Date Egg' Against 'Body Mass (g)' ")  # make a title

# make title centered
fig.update(layout=dict(title=dict(x=0.5)))
# reduce whitespace
fig.update_layout(margin={"r":0,"t":50,"l":0,"b":0})
# show the plot
fig.show()