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

# # Analyse the demographics of the workshop
# 
# Who came, with what level, and how many stayed to the end. 

# In[1]:


import pandas as pd
import matplotlib.pyplot as plt


# Set some default parameters for the matplotlib figures. I want to use the [seaborn](http://seaborn.pydata.org/) style sheet with larger text, all figures to have the same size, and have a white (instead of transparent) background. Also use a higher DPI when saving figures for better quality images.

# In[2]:


plt.style.use(['seaborn', 'seaborn-talk'])
plt.rc('figure', facecolor='#ffffff')
plt.rc('figure', figsize=(10, 5))
plt.rc('figure', dpi=150)


# ## Load the demographics data
# 
# Data from the Google forms sign up. I exported the data (without names and emails) to the file `demographics.csv` in this repository. 
# 
# I'll use `pandas` library to load and manipulate the data.

# In[3]:


data = pd.read_csv('demographics.csv')


# In[4]:


data


# ## Attendance

# First lets see the number of people who signed up:

# In[5]:


len(data)


# From the shared notes, I also know how many people showed up after each day.

# In[6]:


attendance = pd.Series([len(data), 31, 21, 21], index=['Signed up', 'Day 1', 'Day 2', 'Day 3'])


# In[7]:


plt.figure()
ax = plt.subplot(111)
attendance.plot.barh(ax=ax)
ax.invert_yaxis()
ax.set_title('Attendance for the workshop')
ax.set_xlabel('Participants')
plt.tight_layout()
plt.savefig('figures/attendance.jpg')
plt.show()


#  We had a bit of a drop off after the first day. The people who came the seconda day also came on the third. Not too bad.

# ## Programming experience
# 
# Let's first look at the programming experience declared by the participants. I'll keep only the highest level declared and count how many people declared that level.

# In[8]:


# Keep only the last (highest) element in the declared experience
data['highest_experience'] = pd.Series([e.split(', ')[-1] for e in data['experience']])


# In[9]:


plt.figure()
ax = plt.subplot(111)
data.highest_experience.value_counts().plot.barh(ax=ax)
ax.invert_yaxis()
ax.set_title('Highest experience declared')
ax.set_xlabel('Participants')
plt.tight_layout()
plt.savefig('figures/education.jpg')
plt.show()


# The workshop participants have very mixed experience levels, from people who have never programmed before to self-taught programmers. That was expected and it's also not surprising that very few people indicated that they reached their level by taking an advanced course.

# ## Programming languages
# 
# Let's look at the programming language experience. I'll focus on a few main languages and group the others into "Others".

# In[10]:


main_languages = ['Python', 'Matlab/Octave', 'C/C++', 'Fortran', 'R', "None (I told I've never programmed before)"]


# Make a function to mark any other language as "Other" and simplify the "None" entry.

# In[11]:


def only_main(lang):
    if lang not in main_languages:
        lang = 'Other'
    if lang == "None (I told I've never programmed before)":
        lang = 'None'
    return lang


# In[12]:


languages = pd.Series([only_main(lang) 
                       for entry in data['languages'] 
                       for lang in entry.split(', ')]).value_counts()


# In[13]:


languages


# In[14]:


plt.figure()
ax = plt.subplot(111)
languages.plot.barh(ax=ax)
ax.invert_yaxis()
ax.set_title('Programming language experience')
ax.set_xlabel('Participants')
ax.set_xticks(range(0, 22, 3))
plt.tight_layout()
plt.savefig('figures/programming-languages.jpg')
plt.show()


# The dominant language in the department is clearly Matlab, though a few people also know R. Fortran is very low here apparently.

# ## Position at the University
# 
# Finally, we can look at what the participants do.

# In[15]:


plt.figure()
ax = plt.subplot(111)
data['position'].value_counts().plot.barh(ax=ax)
ax.invert_yaxis()
ax.set_title('Position')
ax.set_xlabel('Participants')
ax.set_xticks(range(0, 22, 3))
plt.tight_layout()
plt.savefig('figures/position.jpg')
plt.show()


# Not surprisingly, a majority are grad students. I was very pleased to have someone from the [Na Kupuna program](https://www.hawaii.edu/diversity/seed-programs/na-kupuna-senior-citizen-visitor-program/) and a not insignificant number of faculty.