Skip to content

python_visualization.md

Latest commit

 

History

History
266 lines (219 loc) · 8.67 KB

python_visualization.md

File metadata and controls

266 lines (219 loc) · 8.67 KB

Python visualization

  • visualizing named colors
  • markers
  • starting code 
    %matplotlib inline
# %matplotlib notebook # interactive
import matplotlib.pyplot as plt
import matplotlib as mpl
import matplotlib.cm as cm

1. Create a fully-customizable figure

  • Change the baseline font

    mpl.rcParams['font.family'] = 'sans-serif'
mpl.rcParams['font.sans-serif'] = ['Arial']

  • Create the figure

    fig = plt.figure(figsize = (5, 4))
ax = fig.add_subplot(1, 1, 1)
ax.plot([1, 2, 3, 4], [1, 4, 9, 16], color = 'steelblue', linewidth = 3, marker = 'o', markersize = 10, label = "Quadratic") # add a label to the dataset that will go into a legend

    plt.plot([1, 2, 3, 4], [1, 4, 9, 16])

    fig, (ax1, ax2, ax3) = plt.subplots(3, sharex=True)

  • Label axes

    font_size = 15
ax.set_xlabel("x", fontsize = font_size) # or LaTeX: "$x$"
ax.set_ylabel("f(x)", fontsize = font_size) # or LaTeX: "$f(x)$"

    plt.xlabel("Time With MatPlotLib")
plt.ylabel("How Cool MatPlotLib Seems")

  • Set x tick location and/or labels

    plt.xticks([1, 2, 3], ['','',''])

  • Display legend

    ax.legend(loc='best', frameon=False, fontsize = font_size, markerscale = 1.2)  # loc='lower right'

  • Add a figure label

    ax.text(1, 25, "(A)", fontsize = font_size)

  • Turn off the spines and axes ticks

    for axis in ['bottom','left']:
    ax.spines[axis].set_linewidth(2)
    ax.spines[axis].set_position(("axes", -0.01))
for axis in ['top','right']:
    ax.spines[axis].set_visible(False)

    ax.yaxis.set_ticks_position('left')
ax.xaxis.set_ticks_position('bottom')

  • Add or remove gridlines

    plt.grid()
plt.grid(False)

  • Change the y-scale to log

    ax.set_yscale('log')

  • Set axes ranges

    ax.set_ylim(1, 20)

    plt.xlim((0, 10)) # left, right
plt.ylim(bottom=0, top=100)

  • Set tick label size

    # Change the fontsize of major and minor ticks label 
ax.tick_params(axis='both', which='major', labelsize=10)
ax.tick_params(axis='both', which='minor', labelsize=8)

    plt.tick_params(labelsize=14)

  • Add title of figure

    fig.suptitle('Overall title', fontsize=font_size)
ax.set_title('Axis title', fontsize=font_size)

    plt.title('title')

  • Save the figure

    plt.savefig('quadratic_logscale_half_frame.png', bbox_inches="tight", dpi = 300)

  • Create a figure with pandas

    pandas dataframe plot documentation

    multi_plot = df.plot(kind="bar", figsize=(20,5))
multi_plot.set_xticklabels(df["State"], rotation=45)

2. Quick plots

  • Change the default plot style

      with plt.style.context('ggplot'):
      fig = plt.figure( figsize = (6, 4) )
      ax = fig.add_subplot(1,1,1, facecolor = 'orange') # change background color
      
      # We should add a label to our dataset that will go into a legend
      ax.plot(x, y, label = "Quadratic", color = 'steelblue', linewidth = 3)
      
      # Now we can label the axes. Always label your axes! Who knows what is in the graph otherwise
      ax.set_xlabel("$x$", fontsize = font_size)
      ax.set_ylabel("$f(x)$", fontsize = font_size)
      
      ### Set tick label size
      plt.tick_params(labelsize=14)

      # Display legend
      ax.legend(loc='best', frameon=False, fontsize = font_size-1)

3. Different plot types

  • Documentation

    • Matplotlib enables us to make scatter, bar, histogram, heatmaps, box plots, and violin plots.

    • Simple examples of all of these types at the Matplotlib Gallery page.

  • Histogram

      fig = plt.figure( figsize = (6, 4.5))
  ax = fig.add_subplot(1, 1, 1)

  ax.hist(gpas, bins = 20, color = 'steelblue') # or bins = [0, 2, 3, 4]
  ax.set_xlabel('GPA', fontsize = font_size)
  ax.set_ylabel('Count', fontsize = font_size)
  plt.tick_params(labelsize=12)

  • Histogram as pdf

      fig = plt.figure( figsize = (6, 4.5))
  ax = fig.add_subplot(1, 1, 1)

  ax.hist(gpas, bins = 20, normed = True, color = 'steelblue')
  ax.set_xlabel('GPA', fontsize = font_size)
  ax.set_ylabel('pdf', fontsize = font_size)
  plt.tick_params(labelsize=12)

  • Bar charts

      locations = ['bar', 'grocery store', 'rent', 'clothing', 'solo cups']
  young_spending = [100, 5, 80, 5, 60]
  old_spending = [10, 100, 75, 55, 10]

  fig = plt.figure( figsize = (6, 4))
  ax = plt.subplot(1, 1, 1)

  #For a bar chart we need to create a dummy set of x-data.
  #It's easiest to just use the `range()` function on the number of groups
  num_categories = len(locations)
  x_positions = range(num_categories)
  offset = 0.3 # offset so the bars don't stack on each other

  #Now we can add the 18-24yo
  ax.bar(x_positions, young_spending, offset, label='18-24yo', color='steelblue') # align: 'edge' vs 'center'

  #Then the 25-29yo
  old_xpos = [x_pos + offset +.01 for x_pos in x_positions]
  ax.bar(old_xpos, old_spending, offset, label='25-29yo', color='limegreen')

  #Label it
  ax.set_ylabel('Spending (\$)')
  plt.legend(loc = 'best', fontsize = font_size)

  #Axis labels
  ax.set_xticks([x_pos + offset for x_pos in x_positions])
  ax.set_xticklabels( [label.capitalize() for label in locations], rotation=90, fontsize = font_size)
  ax.set_ylabel('pdf', fontsize = font_size)
  plt.tick_params(labelsize=font_size)

  • Scatter plot

    • Show market size with marker size 
  •   market_size = [200, 100, 400, 20, 50]

  fig = plt.figure(figsize = (5,5))
  ax = fig.add_subplot(1, 1, 1)

  #We add in the x-data points and then the y-data points
  ax.scatter(young_spending, old_spending, s=market_size, color='orange') # or s=5, facecolors='red', edgecolors='black'

  #Labels....
  ax.set_xlabel('18-24yo Spending (\$)', fontsize = font_size)
  ax.set_ylabel('25-29yo Spending (\$)', fontsize = font_size)
  plt.tick_params(labelsize=font_size-2)

    • Show market size with marker color. Mainly three types of color maps:
      • Sequential
      • Diverging
      • Qualitative 
  •   #We get our colormap here
  cm = plt.cm.get_cmap('PuBu')

  #Figure code
  fig = plt.figure(figsize = (5,5))
  ax = fig.add_subplot(1, 1, 1)

  #We add in the x-data points and then the y-data points
  #The scatter function returns itself, we put it into a variable and then feed it into the colorbar function
  sc = ax.scatter(young_spending, old_spending, s=100, c=market_size, cmap=cm, vmin=0, vmax=1000)

  #Labels....
  ax.set_xlabel('18-24yo Spending (\$)', fontsize = font_size)
  ax.set_ylabel('25-29yo Spending (\$)', fontsize = font_size)
  plt.tick_params(labelsize=font_size-2)

  #loop to label data points
  offset = 3
  for index in range(len(locations)):
      print( locations[index], young_spending[index], old_spending[index] )
      plt.text(young_spending[index] + offset, old_spending[index] + offset, \
              locations[index].capitalize())
      
  #plt.text(60, 60, 'Grocery', fontsize = 32)

  #Adding a colorbar
  cbar = plt.colorbar(sc)

  #You need to add a label to the color bar too
  cbar.set_label('Market Size (\$Millions)', fontsize = font_size)

  • Pie chart

      plt.pie(sizes, explode=explode, labels=labels, colors=colors, autopct='%1.1f%%', shadow=True, startangle=140)
  plt.axis('equal') # fits outside plot by square