Python: Bar chart from JSON using Pandas

Pandas for Python will be used.

Pandas will try to make sense of the JSON using several ways so that, in this case, the variations of JSON data used do not need changes in the way that the Pandas calls are made.

This is unlike using just Python where the differences in the JSON require different code statements.

... more to be added ...

To types of data orientations supported by Pandas include the following

Index: dict of row dicts or list of row dicts
Values: list of lists
Columns: dict of column dicts (not used here)

Here are links to the topic series of representing and processing data.

R examples:

Python examples:

Python: Bar chart from text and CSV
Python: Bar chart from JSON
Python: Bar chart from XML
Python: Bar chart from CSV using Pandas
Python: Bar chart from JSON using Pandas (this page)

JavaScript and D3 examples:

JavaScript: SVG bar chart using D3

Julia examples:

Julia: Bar chart from JSON data using StatsPlots

... more to be added ...

Here is the JSON data to be read as a list of lists.

[ ["a",59900], ["b",11302], ["c",14311], ["d",34483], ["e",95247], ["f",17852], ["g",13528], ["h",62509], ["i",46347], ["j",2321], ["k",4449], ["l",25939], ["m",19154], ["n",51311], ["o",54741], ["p",9918], ["q",183], ["r",37468], ["s",44630], ["t",72396], ["u",20651], ["v",7458], ["w",18376], ["x",376], ["y",14814], ["z",234] ]

Here is the Python code [#1]

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
jsonFs1 = "./letters1.json"
print("Loading JSON in \"{0:s}\"".format(jsonFs1))
data1 = pd.read_json(jsonFs1)
fieldNameList1 = ["Letter","Count"]
data1.columns = fieldNameList1
# get first few rows using head
print("")
print("First few rows using head:")
head1 = data1.head()
print("{0:s}".format(str(head1)))
# get column Letter by name Letter
cName1 = "Letter"
print("")
print("Column for {0:s} by name {1:s}:".format(cName1,cName1))
colData1 = data1[[cName1]]
print("{0:s}".format(str(colData1)))
# get colmun Count by index 1
cpos1 = 1
print("")
print("Column for Count by index {0:d}:".format(cpos1))
colData1 = data1.iloc[:,cpos1]
print("{0:s}".format(str(colData1)))
# get row 2 by index 2
rpos1 = 2
rowData1 = data1.iloc[rpos1]
print("")
print("Row {0:d} by index {1:d}:".format(rpos1,rpos1))
print("{0:s}".format(str(rowData1)))
xData1 = data1["Letter"]
yData1 = data1["Count"]
# set plot parameters
fontDict1 = {
	"family" : "Verdana",
	"weight" : "bold",
	"size" : 16
	}
plt.rc("font", **fontDict1)
# make the bar chart
rlen1 = 1
clen1 = 1
width1 = 1024
height1 = 360
dpi1 = 72
figsize1 = (width1/dpi1,height1/dpi1)
fig1,ax1 = plt.subplots(rlen1,clen1,figsize=figsize1,dpi=dpi1)
y_pos1 = np.arange(len(xData1))
plt.bar(y_pos1, yData1)
plt.xticks(y_pos1, xData1)
plt.ylabel("Count")
plt.xlabel("Letter")
plt.title("Count by Letter")
# save the bar chart
print("")
pngFs1 = "./panda10-1-01.png"
print("{0:s} (chart as image)".format(pngFs1))
fig1.savefig(pngFs1,bbox_inches="tight")

Here is the output of the Python code.

Loading JSON in "./letters1.json"
First few rows using head:
  Letter  Count
0      a  59900
1      b  11302
2      c  14311
3      d  34483
4      e  95247
Column for Letter by name Letter:
   Letter
0       a
1       b
2       c
3       d
4       e
5       f
6       g
7       h
8       i
9       j
10      k
11      l
12      m
13      n
14      o
15      p
16      q
17      r
18      s
19      t
20      u
21      v
22      w
23      x
24      y
25      z
Column for Count by index 1:
0     59900
1     11302
2     14311
3     34483
4     95247
5     17852
6     13528
7     62509
8     46347
9      2321
10     4449
11    25939
12    19154
13    51311
14    54741
15     9918
16      183
17    37468
18    44630
19    72396
20    20651
21     7458
22    18376
23      376
24    14814
25      234
Name: Count, dtype: int64
Row 2 by index 2:
Letter        c
Count     14311
Name: 2, dtype: object
./panda10-1-01.png (chart as image)

This is a simple bar chart for the present purposes. Future topics will cover how to make it look nicer.

Here is the JSON data to be read as a list of dicts.

[ {"name":"a", "value":59900}, {"name":"b", "value":11302}, {"name":"c", "value":14311}, {"name":"d", "value":34483}, {"name":"e", "value":95247}, {"name":"f", "value":17852}, {"name":"g", "value":13528}, {"name":"h", "value":62509}, {"name":"i", "value":46347}, {"name":"j", "value":2321}, {"name":"k", "value":4449}, {"name":"l", "value":25939}, {"name":"m", "value":19154}, {"name":"n", "value":51311}, {"name":"o", "value":54741}, {"name":"p", "value":9918}, {"name":"q", "value":183}, {"name":"r", "value":37468}, {"name":"s", "value":44630}, {"name":"t", "value":72396}, {"name":"u", "value":20651}, {"name":"v", "value":7458}, {"name":"w", "value":18376}, {"name":"x", "value":376}, {"name":"y", "value":14814}, {"name":"z", "value":234} ]

Here is the Python code [#2]

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
jsonFs1 = "./letters2.json"
print("Loading JSON in \"{0:s}\"".format(jsonFs1))
data1 = pd.read_json(jsonFs1)
fieldNameList1 = ["Letter","Count"]
data1.columns = fieldNameList1
# get first few rows using head
print("")
print("First few rows using head:")
head1 = data1.head()
print("{0:s}".format(str(head1)))
# get column Letter by name Letter
cName1 = "Letter"
print("")
print("Column for {0:s} by name {1:s}:".format(cName1,cName1))
colData1 = data1[[cName1]]
print("{0:s}".format(str(colData1)))
# get colmun Count by index 1
cpos1 = 1
print("")
print("Column for Count by index {0:d}:".format(cpos1))
colData1 = data1.iloc[:,cpos1]
print("{0:s}".format(str(colData1)))
# get row 2 by index 2
rpos1 = 2
rowData1 = data1.iloc[rpos1]
print("")
print("Row {0:d} by index {1:d}:".format(rpos1,rpos1))
print("{0:s}".format(str(rowData1)))
xData1 = data1["Letter"]
yData1 = data1["Count"]
# set plot parameters
fontDict1 = {
	"family" : "Verdana",
	"weight" : "bold",
	"size" : 16
	}
plt.rc("font", **fontDict1)
# make the bar chart
rlen1 = 1
clen1 = 1
width1 = 1024
height1 = 360
dpi1 = 72
figsize1 = (width1/dpi1,height1/dpi1)
fig1,ax1 = plt.subplots(rlen1,clen1,figsize=figsize1,dpi=dpi1)
y_pos1 = np.arange(len(xData1))
plt.bar(y_pos1, yData1)
plt.xticks(y_pos1, xData1)
plt.ylabel("Count")
plt.xlabel("Letter")
plt.title("Count by Letter")
# save the bar chart
print("")
pngFs1 = "./panda10-2-01.png"
print("{0:s} (chart as image)".format(pngFs1))
fig1.savefig(pngFs1,bbox_inches="tight")

Here is the output of the Python code.

Loading JSON in "./letters2.json"
First few rows using head:
  Letter  Count
0      a  59900
1      b  11302
2      c  14311
3      d  34483
4      e  95247
Column for Letter by name Letter:
   Letter
0       a
1       b
2       c
3       d
4       e
5       f
6       g
7       h
8       i
9       j
10      k
11      l
12      m
13      n
14      o
15      p
16      q
17      r
18      s
19      t
20      u
21      v
22      w
23      x
24      y
25      z
Column for Count by index 1:
0     59900
1     11302
2     14311
3     34483
4     95247
5     17852
6     13528
7     62509
8     46347
9      2321
10     4449
11    25939
12    19154
13    51311
14    54741
15     9918
16      183
17    37468
18    44630
19    72396
20    20651
21     7458
22    18376
23      376
24    14814
25      234
Name: Count, dtype: int64
Row 2 by index 2:
Letter        c
Count     14311
Name: 2, dtype: object
./panda10-2-01.png (chart as image)