Example 1¶
I have written a sript a while ago to plot a long sequence of data. In this first example I plot data represented LIDAR scanning of a road with segments having either cleanded ditches or not, while registering a lot I am intrested here in representing three quantitative variables extracted from LIDAR pointcloud.
What I really want to focus on is the fact that it is possible to split long sequence and plot two subplots,with second representing categories.
In [1]:
import random
import warnings
warnings.filterwarnings('ignore')
import pandas as pd
from matplotlib import pyplot as plt
from math import ceil
ROAD = 16
size_slice = 100
# set heigth ratio using gridspec_kw
width = [1]
heights = [10, 1]
gs_kw = dict(width_ratios=width, height_ratios=heights) # {'width_ratios': [1], 'height_ratios': [5, 1]}
# Helper splits off the df
def chunk(lst, n):
"""Yield successive n-sized chunks from lst.
calling (chunk(list(range(0,15)), 4)))
returns
[[0, 1, 2, 3], [4, 5, 6, 7], [8, 9, 10, 11], [12, 13, 14]]
"""
for i in range(0, len(lst), n):
yield lst[i:i + n]
def roundup(x):
# round up ticks to next hundred
return int(ceil(x / 100.0)) * 100
# parameters
chainage = [9800, 9770, 9700, 9670, 9590, 9480, 9460, 9450, 9440, 9430, 9420, 9410, 9400, 9390, 9380, 9370, 9360, 9350, 9340, 9330, 9320, 9310, 9300, 9290, 9280, 9270, 9260, 9250, 9240, 9230, 9220, 9210, 9200, 9190, 9180, 9170, 9160, 9150, 9140, 9130, 9120, 9110, 9100, 9090, 9080, 9070, 9060, 9050, 9040, 9030, 9020, 9010, 9000, 8990, 8980, 8970, 8960, 8950, 8940, 8930, 8920, 8910, 8900, 8890, 8880, 8870, 8860, 8850, 8840, 8830, 8820, 8810, 8800, 8790, 8780, 8770, 8760, 8750, 8740, 8730, 8720, 8710, 8700, 8690, 8680, 8670, 8660, 8650, 8640, 8630, 8620, 8610, 8600, 8590, 8580, 8570, 8560, 8550, 8540, 8530, 8520, 8510, 8500, 8480, 8470, 8460, 8450, 8440, 8430, 8420, 8410, 8400, 8390, 8380, 8370, 8360, 8350, 8340, 8330, 8320, 8310, 8300, 8290, 8280, 8270, 8260, 8250, 8240, 8230, 8220, 8210, 8200, 8190, 8180, 8170, 8160, 8150, 8140, 8130, 8120, 8110, 8100, 8090, 8080, 8070, 8060, 8050, 8040, 8030, 8020, 8010, 8000, 7990, 7980, 7970, 7960, 7950, 7940, 7930, 7920, 7910, 7900, 7890, 7880, 7870, 7820, 7810, 7800, 7790, 7780, 7770, 7760, 7750, 7740, 7730, 7720, 7710, 7700, 7690, 7680, 7670, 7660, 7650, 7640, 7630, 7620, 7610, 7600, 7590, 7580, 7570, 7560, 7550, 7540, 7530, 7520, 7510, 7500, 7490, 7480, 7470, 7460, 7450, 7440, 7430, 7420, 7410, 7400, 7390, 7360, 7350, 7340, 7330, 7320, 7310, 7300, 7290, 7280, 7270, 7260, 7250, 7240, 7230, 7220, 7210, 7200, 7190, 7180, 7170, 7160, 7150, 7140, 7130, 7120, 7110, 7100, 7090, 7080, 7070, 7060, 7050, 7040, 7030, 7020, 7010, 7000, 6990, 6980, 6970, 6960, 6950, 6940, 6930, 6920, 6910, 6900, 6890, 6330, 6320, 6310, 6300, 6290, 6280, 6270, 6260, 6250, 6240, 6230, 6220, 6210, 6200, 6190, 6180, 6170, 6160, 6150, 6140, 6130, 6120, 6110, 6100, 6090, 6080, 6070, 6060, 6050, 6040, 6030, 6020, 6010, 6000, 5990, 5980, 5970]
sci = [81, 42, 160, 113, 71, 83, 65, 110, 136, 109, 133, 189, 196, 189, 179, 168, 174, 180, 181, 202, 140, 143, 105, 136, 194, 137, 157, 205, 295, 287, 249, 224, 269, 234, 141, 164, 189, 147, 146, 133, 173, 163, 168, 188, 178, 159, 184, 191, 157, 144, 181, 152, 164, 161, 145, 193, 217, 211, 189, 196, 187, 171, 173, 162, 157, 143, 148, 158, 187, 152, 170, 200, 204, 155, 157, 179, 202, 202, 206, 216, 212, 206, 167, 139, 194, 152, 158, 151, 174, 181, 158, 166, 161, 175, 188, 174, 194, 249, 201, 180, 163, 152, 133, 77, 99, 128, 144, 134, 121, 153, 180, 196, 177, 168, 177, 194, 185, 269, 247, 255, 263, 285, 282, 211, 190, 201, 205, 217, 202, 180, 178, 232, 201, 190, 227, 178, 139, 172, 208, 210, 190, 179, 156, 169, 182, 196, 173, 154, 180, 184, 206, 182, 239, 237, 246, 207, 192, 213, 227, 200, 201, 187, 212, 255, 231, 218, 221, 203, 242, 271, 213, 186, 185, 177, 214, 222, 205, 182, 161, 210, 303, 248, 196, 236, 279, 222, 264, 187, 243, 209, 167, 177, 196, 191, 209, 166, 240, 178, 234, 222, 222, 207, 279, 291, 302, 225, 210, 289, 313, 317, 270, 232, 188, 174, 198, 214, 187, 196, 157, 201, 227, 253, 233, 229, 299, 269, 233, 235, 261, 276, 269, 236, 265, 301, 235, 221, 239, 206, 182, 164, 207, 224, 206, 182, 214, 183, 185, 244, 242, 267, 299, 293, 275, 248, 219, 259, 210, 175, 246, 286, 254, 225, 201, 169, 207, 228, 184, 185, 190, 189, 161, 124, 133, 139, 155, 154, 136, 141, 132, 103, 88, 91, 94, 120, 140, 164, 122, 144, 136, 126, 133, 165, 135, 145]
bci = [67, 57, 113, 71, 74, 75, 61, 95, 106, 104, 113, 125, 121, 124, 136, 126, 138, 136, 127, 133, 104, 112, 91, 98, 162, 108, 131, 144, 146, 145, 133, 138, 157, 156, 106, 122, 129, 103, 123, 99, 133, 119, 122, 129, 122, 111, 116, 120, 113, 103, 122, 117, 121, 110, 100, 116, 116, 120, 112, 100, 101, 97, 94, 91, 97, 75, 88, 99, 125, 102, 114, 111, 114, 91, 92, 125, 123, 140, 141, 153, 138, 130, 114, 114, 129, 96, 108, 117, 117, 125, 124, 101, 105, 118, 100, 104, 119, 139, 154, 130, 129, 123, 126, 80, 97, 116, 103, 95, 97, 111, 124, 124, 146, 125, 121, 138, 122, 167, 180, 176, 190, 176, 184, 147, 147, 153, 161, 157, 148, 135, 127, 146, 138, 126, 165, 141, 123, 135, 149, 144, 125, 131, 105, 102, 124, 139, 121, 134, 137, 122, 127, 130, 144, 152, 146, 130, 130, 126, 144, 137, 127, 124, 132, 143, 142, 155, 138, 147, 151, 148, 152, 149, 137, 136, 158, 154, 144, 142, 140, 149, 174, 180, 158, 176, 181, 164, 135, 142, 166, 167, 146, 145, 161, 162, 166, 125, 155, 135, 145, 152, 163, 156, 173, 198, 176, 154, 150, 167, 191, 192, 168, 148, 137, 117, 126, 143, 139, 130, 125, 146, 149, 162, 188, 172, 178, 171, 144, 159, 161, 163, 176, 152, 157, 167, 154, 143, 154, 136, 120, 100, 113, 139, 140, 128, 136, 107, 129, 157, 139, 151, 165, 167, 161, 167, 143, 140, 121, 164, 185, 190, 179, 160, 144, 131, 160, 176, 144, 150, 139, 145, 99, 85, 92, 109, 113, 110, 108, 107, 105, 92, 63, 100, 76, 81, 99, 120, 93, 109, 98, 87, 129, 107, 86, 100]
bdi = [164, 154, 255, 143, 181, 116, 148, 147, 157, 212, 94, 191, 127, 177, 216, 185, 140, 144, 169, 147, 124, 195, 141, 147, 158, 145, 138, 146, 75, 108, 112, 108, 120, 156, 118, 151, 160, 115, 74, 123, 167, 137, 128, 114, 133, 114, 85, 96, 147, 136, 128, 142, 176, 103, 105, 96, 45, 88, 47, 32, 105, 108, 69, 100, 113, 65, 38, 95, 129, 87, 116, 40, 103, 33, 94, 156, 44, 150, 182, 210, 134, 112, 109, 199, 165, 34, 111, 152, 135, 162, 154, 40, 114, 107, 86, 39, 115, 80, 205, 193, 117, 189, 225, 186, 165, 171, 143, 145, 157, 120, 163, 135, 187, 151, 126, 188, 116, 208, 185, 187, 244, 158, 191, 123, 161, 154, 178, 174, 160, 126, 111, 159, 109, 169, 209, 158, 158, 177, 161, 148, 122, 166, 48, 103, 157, 151, 109, 181, 94, 57, 94, 101, 147, 137, 135, 58, 134, 96, 125, 161, 144, 126, 157, 141, 129, 210, 144, 173, 132, 135, 183, 221, 167, 157, 216, 170, 150, 149, 188, 185, 153, 227, 149, 173, 169, 185, 169, 176, 173, 182, 206, 200, 253, 192, 93, 205, 185, 164, 121, 145, 205, 194, 204, 264, 208, 136, 173, 198, 184, 186, 162, 152, 160, 145, 156, 189, 181, 154, 174, 202, 162, 182, 97, 228, 174, 192, 172, 205, 182, 154, 182, 158, 191, 167, 165, 166, 179, 145, 59, 74, 85, 153, 172, 153, 139, 122, 130, 221, 142, 173, 171, 141, 158, 154, 147, 133, 124, 302, 260, 220, 237, 216, 196, 168, 184, 257, 235, 233, 191, 211, 138, 116, 145, 98, 177, 167, 169, 157, 184, 96, 91, 105, 159, 101, 148, 190, 152, 162, 159, 119, 99, 132, 120, 131]
"""
sci = [random.randint(40,320) for i in range(0,294)]
bci = [random.randint(50,200) for i in range(0,294)]
bdi = [random.randint(30,310) for i in range(0,294)]
"""
# height is the value for control (not influenced) in the lowest subplot
height = [1.0, 1.0, 1.0, float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), 1.0, 1.0, float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, float('nan'), 1.0, 1.0, 1.0, float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, float('nan'), float('nan'), float('nan'), float('nan'), 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), 1.0, 1.0, 1.0, float('nan'), float('nan'), float('nan'), 1.0, 1.0, 1.0, 1.0, float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, float('nan'), 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, float('nan'), float('nan'), float('nan'), 1.0, 1.0, 1.0, float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), 1.0, float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), 1.0, float('nan'), float('nan'), 1.0, 1.0, float('nan'), float('nan'), float('nan'), 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, float('nan'), float('nan'), float('nan'), float('nan'), float('nan'), 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, float('nan'), float('nan')]
# height_acm = list(df2_ac_mixed.height)
height_acm = [3.0
if i!=1.0 else float('nan')
for i in height]
height_ac = [random.choice([2.0, float('nan')])
if i==3.0 else float('nan')
for i in height_acm]
# notice here start and end computed based on chainage
start_of_chain, end_of_chain = min(chainage), max(chainage)
chainage_range = range(roundup(start_of_chain), roundup(end_of_chain), 100)
In [2]:
# parameters decorated
_args = [chainage,sci,bci, bdi, height, height_ac, height_acm]
# divide all iterables into x-sized slices
_par = [(chunk(i,size_slice)) for i in _args]
In [3]:
zipped_slices2 = zip(*_par)
def the_plotter(num_figures, value_collection):
# chainage, sci, bci, bdi, marker_chain, height,
# marker_chain_ac, height_ac, marker_chain_ac_mixed, height_acm
chainage = value_collection[0]
sci = value_collection[1]
bci = value_collection[2]
bdi = value_collection[3]
height = value_collection[4]
height_ac = value_collection[5]
height_acm = value_collection[6]
f, (ax1, ax2) = plt.subplots(2, 1, constrained_layout=True,
sharey=False, figsize=(40, 15),
gridspec_kw=gs_kw)
ax1.plot(chainage, sci, color="black", linewidth= 2.0)
ax1.plot(chainage, bci, color="magenta")
ax1.plot(chainage, bdi, color="dodgerblue") # indigo
ax1.set_ylim(top=500)
ax1.legend(('sci', 'bci', 'bdi'), loc=2, prop={'size': 20}) # legend size set
# it is a bit countre intuitive
# you do not split chainage_range into
# smaller sequences since it is longer than
# visualized sequence it is cropped outside of it
ax1.set_xticks(list(chainage_range))
ax1.set_xticklabels(list(chainage_range), fontsize=14) # list(xticks_computation(chainage_range)
# ===================================================
# PLOT 2
# control
ax2.plot(chainage, height, linestyle='None', color="limegreen", marker='s', markersize=6)
# 2x factor match
ax2.plot(chainage, height_ac, linestyle='None', color="red", marker='s', markersize=6)
# is invent & any pms
ax2.plot(chainage, height_acm, linestyle='None', color="gold", marker='s', markersize=6)
ax2.yaxis.set_ticks([1,2,3])
ax2.set_yticklabels(['control','2x factor match','is invent & any/none pms'], fontsize=14)
ax2.set_ylabel('factor levels', fontsize = 16)
# =========================================================
# BARS
# add a bar chart to the upper subplot
width = 10 # the width of the bars, approx(end_of_chain - start_of_chain)/300
x = chainage
y = [i*500 for i in height if i]
y2 = [i*250 for i in height_ac if i]
y3 = [i*166 for i in height_acm]
# control
rects1 = ax1.bar(x, y, width, color='limegreen', alpha = 0.6)
# is invent & any/none pms
rects3 = ax1.bar(x, y3, width, color='gold', alpha = 0.6)
# 2x factor match
rects2 = ax1.bar(x, y2, width, color='red', alpha = 0.3)
# plt.savefig(f"roadnumber{ROAD}part{num_figures}_slice2000_v20200507.png")
# TO DO refactor as main
counter1=0
for chunk_element in reversed(list(zipped_slices2)):
counter1 += 1
the_plotter(counter1, chunk_element)
Example 2 The stacked barplot¶
This one is about representing distribution of frequencies (in %) of categorical variables in a correspondence analysis. This one is pulled from my favorite tuto about CA.
To create a cumulative barplot for percentages of a cross table I follow these steps:
Calculate the percentage for each category Calculate the cumulative percentage for each category Create a barplot with the categories on the x-axis and the cumulative percentages on the y-axis. you can add the percentage for each category as labels on the bars or as a separate table next to the plot
https://www.geeksforgeeks.org/create-a-stacked-bar-plot-in-matplotlib/
In [67]:
import pandas as pd
import numpy as np
data_ca = r'C:\thisAKcode.github.io\Pelican\content\other\CSV.csv'
df = pd.read_csv(data_ca, encoding='utf-8')
crosstab_orig = df.iloc[1:9, 1:-1].to_numpy()
countries = df.iloc[1:-1, 0].tolist()
prize_categories = df.columns[1:-1].tolist()
df
Out[67]:
In [68]:
# x.j
row_totals = np.sum(crosstab_orig, axis=1)
# xi.
column_totals = np.sum(crosstab_orig, axis=0)
# x..
grand_total = np.sum(crosstab_orig)
the code below will generate a stacked bar plot showing the percentages across rows for each category.
In [69]:
from matplotlib import pyplot as plt
data = {
'col0': ['Country', 'Germany', 'Canada', 'France', 'UK', 'Italy', 'Japan', 'Russia', 'US', 'Total'],
'Chemistry': [None, 24.0, 4.0, 8.0, 23.0, 1.0, 6.0, 4.0, 51.0, 121.0],
'Economic sciences': [None, 1.0, 3.0, 3.0, 6.0, 1.0, 0.0, 3.0, 43.0, 60.0],
'Literature': [None, 8.0, 2.0, 11.0, 7.0, 6.0, 2.0, 5.0, 8.0, 49.0],
'Medicin': [None, 18.0, 4.0, 12.0, 26.0, 5.0, 3.0, 2.0, 70.0, 140.0],
'Peace': [None, 5.0, 1.0, 10.0, 11.0, 1.0, 1.0, 3.0, 19.0, 51.0],
'Physics': [None, 24.0, 4.0, 9.0, 20.0, 5.0, 11.0, 10.0, 66.0, 149.0]
}
df = pd.DataFrame(data)
df_percent = df.iloc[1:9, 1:8].div(df.iloc[1:9, 1:8].sum(axis=1), axis=0) * 100
ax = df_percent.plot(kind='bar', stacked=True)
# Set the x-axis labels
ax.set_xticklabels(df['col0'][1:9])
# Set the y-axis label
ax.set_ylabel('Percentage')
# Set the chart title
ax.set_title('Stacked Bar Plot of Percentages Across Rows')
# Display the legend
ax.legend(loc=(1.1, 0.5))
# Show the plot
plt.show()
In [70]:
df
Out[70]:
In [71]:
df_percent
Out[71]:
To create a stacked bar plot of percentages across columns with prize categories on the x-axis and country percentages on the y-axis, you can follow these steps:
In [72]:
import pandas as pd
# Transpose the DataFrame
df_transposed = df.set_index('col0').transpose()
# Recompute sum at the last line (Total) grouped by prize categories
df_transposed['Total'] = df_transposed.sum(axis=1)
In [73]:
df_transposed # .drop(columns = 'Country', inplace=True)
df_transposed2 = df_transposed.drop(columns = 'Country', inplace = False)
df_transposed2
Out[73]:
In [74]:
df_transposed2.drop(columns=df_transposed2.columns[-1], inplace=True)
df = df_transposed2
df.loc['Total'] = df.sum()
# df = df.drop([0, 9])
df_percent = df.iloc[0:6, 0:9].div(df.iloc[0:6, 0:9].sum(axis=1), axis=0) * 100
# 19 18
ax = df_percent.plot(kind='bar', stacked=True)
# Set the x-axis label
ax.set_xlabel('Prize Categories')
# Set the y-axis label
ax.set_ylabel('Percentage')
# Set the chart title
ax.set_title('Stacked Bar Plot of Percentages Across Columns')
# Display the legend
ax.legend(loc=(1.1, 0.5))
# Show the plot
plt.show()
In [75]:
df
Out[75]:
In [76]:
df_transposed
Out[76]:
In [77]:
df_percent
Out[77]:
In [ ]: