How to plot [number of items, price sold] as histogram graph in gnuplot

Question

I have a csv file that prices of items and the number of items for that price. For example a .csv file:

32,4.23
980,44.53
533,4.23

The above csv is interpreted like this: 32 items are available for $4.23 and 980 items are available for $44.53. The data is not sorted (see the 3rd line in the .csv - it has $4.23 twice).

Here is the data of the .csv file:

count,cost
897,4.19
903,8.28
916,5.19
923,7.79
935,7.40
936,7.53
952,2.00
963,7.30
966,10.67
980,7.58
987,11.14
993,18.43
997,72.60
1005,6.52
1007,12.91
1015,11.83
1022,5.55
1056,6.39
1058,5.30
1063,5.09
192,34.01
1067,4.55
1067,14.01
1078,5.54
3236,64.36
1095,4.58
1101,5.18
1106,5.53
100,135.25
1114,8.61
1115,8.06
1116,5.34
1116,5.86
1119,5.28
1009,3.38
1122,5.99
1015,36.20
204,78.03
1143,10.84
1145,5.86
1148,7.83
1155,18.06
1159,6.53
1170,2.56
1173,9.02
1185,12.86
1191,7.84
1191,8.08
1203,9.07
1232,4.15
1247,111.84
1256,6.93
1132,4.64
1266,8.53
1271,7.53
23060,24.67
1285,14.90
1289,3.44
235,28.46
1311,7.19
1325,7.21
1345,10.31
1364,9.74
1374,4.84
1380,6.38
1243,36.20
1386,7.52
1389,6.15
1403,3.49
1408,468.95
1414,4.25
1421,6.77
1423,4.91
258,39.47
1434,8.55
1437,6.08
1441,4.71
1448,4.59
1456,4.86
1490,14.12
1502,3.83
1504,213.23
1515,18.25
1365,4.86
1521,16.25
1522,10.25
4580,57.18
1537,4.51
1543,7.68
1550,276.50
1553,11.13
1572,6.46
289,72.79
1606,51.78
1615,7.50
1622,356.00
1654,4.53
1663,5.90
1675,7.69
1680,16.35
1682,5.49
1684,16.64
1697,10.40
1715,4.64
1716,3.79
334,57.40
1757,6.63
322,51.86
1793,13.33
1793,9.21
324,72.16
1843,13.43
1869,7.45
1878,3.69
1891,3.65
1709,60.85
1906,14.54
173,137.71
1936,6.75
371,58.96
1950,11.09
1766,62.87
1968,4.21
1993,8.76
1995,6.40
2026,8.29
2029,9.45
2031,9.96
1843,3.36
392,50.71
2072,4.10
2084,292.42
2089,7.86
1885,4.07
2099,6.28
2108,6.17
2113,412.11
2118,9.99
2118,187.66
2124,4.52
2130,5.92
193,142.42
2164,7.04
2172,6.81
2191,7.46
2206,7.76
1988,35.94
2266,5.57
2271,8.79
2283,5.87
2294,7.80
2307,17.15
2310,7.02
2318,4.80
2329,9.95
2354,7.45
2386,7.21
2150,36.04
2411,9.32
2446,7.59
2468,6.35
224,87.99
2248,76.16
229,132.84
2563,3.36
488,51.87
2310,62.56
2570,8.19
2589,11.51
2594,9.23
2600,7.03
2600,4.02
2614,6.33
2618,4.28
2653,12.36
2388,3.11
7982,28.13
511,52.48
2772,54.07
2789,9.15
2811,12.07
535,53.91
2863,2.99
2869,7.74
517,39.44
2919,9.96
2959,31.91
2983,8.55
2685,58.88
2987,12.82
2996,8.47
3006,13.38
2723,72.85
2728,3.97
3066,7.72
553,49.95
3084,11.93
3090,6.83
3092,4.53
3104,151.76
3130,11.39
3130,5.81
597,50.38
3140,8.71
3154,8.07
2876,44.07
3212,11.66
3218,4.81
580,23.74
3237,3.72
583,22.80
3254,15.25
3260,6.03
3262,54.73
3263,7.75
3273,9.31
3341,691.45
3356,351.09
3067,3.59
3412,10.82
3110,45.15
3497,9.98
3498,279.80
317,96.84
320,132.21
3574,210.91
3605,2.49
3611,9.61
3660,5.36
3674,7.61
332,93.87
3706,12.42
3740,1.87
3818,9.25
3852,4.73
349,105.21
4116,17.27
4129,10.28
4143,665.57
376,118.20
4216,10.98
4241,10.39
4249,6.51
4302,24.39
3878,45.22
389,140.03
4331,9.61
4354,8.89
4357,7.07
4357,80.19
4433,462.20
4573,4.84
4621,445.18
4641,13.52
4724,5.36
4742,10.59
4751,8.18
4282,53.04
877,24.44
4880,8.77
4934,257.99
4477,2.71
5066,7.00
941,32.10
5234,6.90
5256,10.07
5326,6.33
1023,51.49
5395,7.70
488,122.44
5524,10.54
5613,9.56
5748,6.59
5841,3.71
5866,19.48
5925,9.48
5997,12.19
6023,13.34
6027,2.16
6048,6.70
1169,61.96
6158,5.68
6200,18.95
6231,9.08
6231,4.03
6248,7.50
6278,2.70
6588,9.40
1190,73.63
6621,1.99
6680,7.70
6740,6.27
6779,6.51
6821,8.32
6890,11.28
1241,26.32
6919,520.20
7007,10.83
7093,4.50
7119,97.28
1300,19.18
7370,6.33
7391,10.07
7394,9.03
7559,5.81
7636,6.17
7764,4.97
7878,505.64
7900,5.54
7918,5.98
7945,6.90
1459,26.65
8107,470.95
8219,8.30
8292,8.80
8460,5.78
8470,320.78
8516,11.20
8646,3.85
8654,13.42
8663,7.49
8682,14.13
8840,247.85
8889,6.91
8971,5.38
9033,7.79
9041,4.11
9057,6.39
9203,28.90
9276,14.13
1726,41.63
8659,62.79
1771,23.59
9865,55.14
9895,12.81
9955,3.71
9964,8.23
10070,10.79
10072,3.78
10087,6.17
10255,6.12
10277,11.25
10309,7.16
10382,4.43
10419,6.96
10711,13.98
10779,4.34
10806,4.05
10906,10.24
10941,11.39
10951,2.92
11064,2.65
2111,53.94
11115,6.21
11151,174.18
11232,239.23
11234,19.92
11362,130.80
11413,4.10
2059,51.61
11559,4.34
11614,3.77
11626,16.75
11644,3.46
11817,3.30
11835,7.01
12005,2.83
12203,6.35
12249,5.37
12305,11.54
12388,646.66
12589,567.88
12870,6.88
12950,7.56
13014,13.51
13084,4.93
13110,9.09
13257,8.14
12115,38.85
13720,4.49
13816,6.49
13962,79.15
13999,4.87
14047,6.04
1320,131.86
2667,33.61
14875,64.52
14951,277.46
15130,673.37
15184,7.26
15282,2.49
15306,10.88
13800,54.71
15402,7.20
15424,4.75
15513,5.46
15730,7.96
15795,5.66
15997,5.19
14434,40.45
16104,8.94
3066,50.96
16410,84.17
16452,4.36
16597,8.99
16832,9.73
16925,3.69
3080,14.10
15511,38.47
17303,4.79
17322,5.42
15784,56.67
17630,2.12
17955,5.98
18035,8.73
3248,38.16
18224,263.03
16450,4.13
18965,6.55
19104,4.89
19213,11.09
19213,5.16
17315,3.09
19242,6.61
19277,5.57
1736,159.14
19292,674.73
19510,283.02
19564,8.34
19801,19.79
19835,13.09
19924,8.45
3597,14.79
20069,8.21
3703,46.30
21472,4.71
21538,686.42
21933,11.40
21999,9.05
22521,13.24
22611,7.75
22996,380.99
23196,13.91
23423,11.52
24182,8.17
24675,1.69
24718,2.80
25020,11.53
25145,16.15
25156,2.67
25196,688.55
25384,2.40
25400,5.89
25583,3.12
26009,7.74
26160,14.12
26951,4.94
24561,3.69
27409,2.49
27567,11.40
27623,5.96
25064,39.24
27983,11.80
28144,4.27
28542,7.48
29030,3.60
2681,157.99
2686,146.40
31009,15.01
31290,5.29
34030,66.96
34562,7.69
34928,692.10
35308,16.59
35860,7.86
36718,10.07
37577,5.34
38671,10.01
38917,8.56
39880,2.43
40084,3.93
42657,3.69
43579,9.70
44358,3.18
46264,3.74
48585,13.95
49628,14.05
50473,12.37
52322,2.45
52821,2.94
55972,7.46
56335,13.83
58709,8.75
58943,12.26
59606,8.88
60101,5.43
11449,51.99
63321,7.40
63530,13.77
66155,6.07
72937,7.30
73588,2.84
73833,2.97
74155,1.99
76127,3.27
76976,2.68
77694,11.60
83614,12.99
84766,12.39
84814,15.68
85046,6.38
85791,6.03
86881,11.66
87195,2.87
87507,2.52
88534,4.29
93611,8.21
95466,2.66
96969,13.11
97697,2.22
99125,9.10
101281,3.74
103736,3.18
107233,2.58
107524,12.50
126053,9.13
129159,12.95
135945,14.39
137074,5.68
140575,9.58
150164,11.10
179075,9.46
184129,6.86
234357,3.17
288633,3.61
310993,2.70
330076,4.80
335276,3.54
372560,5.03
443936,6.20
468026,4.53

Here is the code that I tried:

DATAFILE = "/tmp/3.csv"

set terminal pngcairo truecolor
set output "a.png"

set grid

set xtic #rotate by 90

set style fill solid 0.66

set boxwidth 0.50
set datafile separator ","

# Define a function that maps a number to a bin:
#  startx, endx: expected interval of values for x
#  n: number of bins
bin(x, startx, endx, n) = (x > endx)? n - 1 : ((x < startx)? 0 : floor(n * ((x - startx) / (endx - startx))))

# Define a function to map bin back to a real value
start_of_bin(i, startx, endx, n) = startx + i * ((endx - startx) / n) 

N = 24 # number of bins
START = 0.0 # start of range (we are interested into)
END = 12.0 # end of range

# Configure x-axis
set xrange [0:N]
set for [i=0:N:+5] xtics (sprintf("%.1f", start_of_bin(i, START, END, N)) i)

# Plot histogram: map (multiple times) every bin to 1.0. 
# Must use smooth freq to actually count frequencies (see `help smooth freq`) !!
plot DATAFILE using (bin($2, START, END, N)):(1.0) smooth freq with boxes title "Some distribution"

Here is what I get:

I want to be able to see what is the most popular price (example: are most items are sold for between $4-$4.5). It like these count items need to be summed??

How can I plot a histogram for the above data (do I need preprocessing?).

Another possibility is something like https://stackoverflow.com/a/7454274/881362 but that does not appear to be a histogram because it does not have bins.

I am a bit in knots and need some help with creating some clarity.

This site is not a discussion forum. **DO NOT** post images of code, links to code, data, error messages, etc. - copy or type the text into the question. Tak the time to prepare a clear concise question that includes a complete [Minimal Reproducible Example](https://stackoverflow.com/help/minimal-reproducible-example "Minimal Reproducible Example") containing both data, and code that can be copied and pasted along with a clear description of what is wrong with your current effort. — itprorh66, Sep 17 '22 at 16:20
@itprorh66 I don't think OP is having a discussion forum, nor posting images of code or links. It's clearly asked `How can I plot a histogram for the above data (do I need preprocessing?).` and did some research (perhaps not enough) — , Sep 17 '22 at 16:27
@Trewq please update the post by posting a link to a sample csv file. Thank you. — , Sep 17 '22 at 16:29
what did you try? Where is your code? At least you could create code which loads data or create code with data directly in code. And you could display it in any way. — furas, Sep 17 '22 at 20:30
with pandas I need only `df['cost'].hist()` to create any histogram. But it may need to use `groupby` to create bins and sum `count`. And your link also shows histogram - but probably it uses default value for `bins` and you don't see any `bins` in code. — furas, Sep 17 '22 at 20:34

theozh · Accepted Answer · 2022-09-18T08:14:15.653

You can easily do it with gnuplot (basically 2 lines, the rest is formatting of the graph). I don't know where you got your code from and I also don't know what this data and numbers are. Apparently, your prices have a wide range from 1 to 700. You can define bins (here: myBinwidth=0.5). Since the sums in the bins vary over several orders of magnitude you better use logscale y to get an overview about your data. If you want to zoom-in for the maximum counts (e.g. set xrange [0:20]), check the lower plot. You will influence the shape of the histogram with the choice of your binwidth.

Edit: you might want to shift the histogram bars by half of myBinwidth, then the bar, e.g. for price >=2.5to <3.0 is ranging from 2.5 to 3.0 and not as perviously centered at 2.5

Script:

### histogram
reset session

FILE = "SO73755867.dat"

set datafile separator comma

myBinwidth = 0.5
bin(x)     = floor(x/myBinwidth)*myBinwidth

set boxwidth myBinwidth 
set logscale y
set style fill solid 0.5
set grid x,y
set tics out

set multiplot layout 2,1

    plot FILE u (bin($2)+myBinwidth/2.):1 smooth freq w boxes

    unset logscale
    set xtic 1
    set mxtics 2
    set xrange[0:30]
    replot
unset multiplot
### end of script

Result:

furas · Answer 2 · 2022-09-18T01:07:52.957

Using pandas I would first use cut() to create bins

df['bin'] = pd.cut(df['cost'], bins=10)

I can also use list with ranges bins=[0, 4, 4.53, 10]

df['bin'] = pd.cut(df['cost'], bins=[0, 4, 4.53, 10])

    count  cost           bin
0  288633  3.61    (0.0, 4.0]
1  310993  2.70    (0.0, 4.0]
2  330076  4.80  (4.53, 10.0]
3  335276  3.54    (0.0, 4.0]
4  372560  5.03  (4.53, 10.0]
5  443936  6.20  (4.53, 10.0]
6  468026  4.53   (4.0, 4.53]

Next I would use groupby() to group the same bins and run sum('count')

new_df = df.groupby('bin').sum('count')

                count   cost
bin                         
(0.0, 4.0]     934902   9.85
(4.0, 4.53]    468026   4.53
(4.53, 10.0]  1146572  16.03

And finally I would plot it as bars

new_df['count'].plot(kind='bar', rot=0)  # don't rotate labels

Full example code (with small changes):

data = '''count,cost
288633,3.61
310993,2.70
330076,4.80
335276,3.54
372560,5.03
443936,6.20
468026,4.53'''

import pandas as pd
import io
import matplotlib.pyplot as plt

#df = pd.read_csv(filename_or_url)
#df = pd.read_csv('https://pastebin.com/raw/pneEpa7M')
df = pd.read_csv(io.StringIO(data))
print(df)

min_price = df['cost'].min()
max_price = df['cost'].max()
print('min:', min_price)
print('max:', max_price)

#df['bin'] = pd.cut(df['cost'], bins=10)
#df['bin'] = pd.cut(df['cost'], bins=[0, 4, 4.53, 10])
df['bin'] = pd.cut(df['cost'], bins=[min_price, 4, 4.53, max_price])
print(df)

new_df = df.groupby('bin').sum('count')
print(new_df)

new_df['count'].plot(kind='bar', rot=0)
plt.show()

Pandas Doc: cut, groupby, hist

Similar question: python - How to draw a histogram using pandas cut - Stack Overflow

thanks! So I do need to do preprocessing and do not have a clean way of doing this directly from gnuplot it seems — Trewq, Sep 17 '22 at 21:12
I don't know `gnuplot` commands - it may need some complex function for this. — furas, Sep 17 '22 at 21:21

How to plot [number of items, price sold] as histogram graph in gnuplot

2 Answers2