如何创建散点图以将 2 个嵌套字典与另一个具有相同键的嵌套字典进行比较

Question

我有三个嵌套字典：data_geo1、data_geo2、data_ali，它们都由两个键和一个值组成。目标是相互比较这些值。字典的第一个键是一样的，但第二个，其中 30 个与第一个字典共享，其中 20 个与第二个字典共享！不幸的是，键的方向彼此不同。目标是使用 matplotlib 和 numpy 数组创建散点图，以将第一个和第二个字典与 data_ali 中的数据进行比较。有点复杂，不知道从何说起！下面是我写的用于创建字典的内容：

import os
import numpy as np

path = "/home/ali/Desktop/data/"
root = "/home/ali/Desktop/SAMPLES/"



data_geo1={}
with open(path+"GSE98212_H_DE_genes_count.txt","rt") as fin: #data for sample 1-30
    h = fin.readline()
    sample1 = h.split()
    sample_names = [s.strip('"') for s in sample1[1:31]]
    for l in fin.readlines():
        l = l.strip().split()
        if l:
            gene= l[0].strip('"')
            data_geo1[gene] = {}
            for i, x in enumerate(l[1:31]):
                data_geo1[gene][sample_names[i]] = int(x)

#print(data_geo1)

data_geo2={}
with open (path+"GSE98212_L_DE_genes_count.txt","rt") as fin:
        h= fin.readline()
        sample2=h.split()
        sample_names=sample2[1:21]
        for l in fin.readlines():
            l = l.strip().split()
            if l:
                gene= l[0].strip()

            data_geo2[gene]={}
            for i,x in enumerate (l[1:21]):
                data_geo2[gene][sample_names[i]]= int(x)

#print(data_geo2)

data_ali={}

for sample_name in os.listdir(root):
    with open(os.path.join(root, sample_name, "counts.txt"), "r") as fin:
        for line in fin.readlines():
            gene, reads = line.split()
            reads = int(reads)
            if gene.startswith('ENSG'):
                data_ali.setdefault(gene, {})[sample_name] = reads

#print(data_ali)

每个字典结构的例子：

data_geo1：{'ENSG00000110514'：{'Sample_19-leish_023_v2'：709，'Sample_4-leish_012_v3'：501，'Sample_25-leish027_v2'：690，'Sample_6-leish_015_v3'：27070，'Sample_6-leish_015_v3'：27070 , 'Sample_20-leish_023_v3': 619, 'Sample_18-leish_022_v3': 678, 'Sample_10-leish_017_v3': 477, 'Sample_13-leish_019_v2': 460, 'Sample_1-Leish_011_v2': 574, 'Sample_11-leish_018_v2': 566, ' Sample_3-leish_012_v2': 632, 'Sample_2-leish_011_v3': 388, 'Sample_29-leish032_v2': 661, 'Sample_8-leish_016_v3': 372, 'Sample_28-leish028_v3': 533, 'Sample_27-leish028_v2': 582, 'Sample_26- leish027_v3': 624, 'Sample_12-leish_018_v3': 653, 'Sample_5-leish_015_v2': 421, 'Sample_16-leish_021_v3': 376, 'Sample_21-leish_024_v2': 668, 'Sample_9-leish_017_v2': 583, 'Sample_24-leish026_v3' ：590，'sampe_22-leish_024_v3'：537，'sample_14-leish_019_v3'：438，'sample_30-leish032_v3'：494，'sample_7-leish_016_v2'：518，sample_15-leish_15-leish_15-leish_17.17; }

data_geo2:{'ENSG00000110514': {'Sample_19': 518, 'Sample_10': 468, 'Sample_20': 517, 'Sample_9': 431, 'Sample_8': 522, 'Sample_7': 437, 'Sample_6' ：491，'Sample_5'：461，'Sample_4'：442，'Sample_3'：667，'Sample_2'：438，'Sample_1'：378，'Sample_14'：345，'Sample_13'：424，'Sample_18'：570 , 'Sample_15': 492, 'Sample_16': 486, 'Sample_12': 401, 'Sample_17': 489, 'Sample_11': 464}

data_ali: 'ENSG00000110514': {'Sample_19-leish_023_v2': 710, 'Sample_16-leish_021_v3': 380, 'Sample_20': 517, 'Sample_24-leish026_v3': 593, 'Sample_16-leish_01': 593, 'Sample_16-leish_01' -leish_018_v3'：661，'sample_22-leish_024_v3'：539，'sample_23-leish026_v2'：710，'sample_25-leish027_v2'：689：689，'sample_18-leish_18-leish_0222_v3'sample_________3'：681'：681，14.14'' 394, 'Sample_13-leish_019_v2': 464, 'Sample_1-Leish_011_v2': 574, 'Sample_11-leish_018_v2': 571, 'Sample_20-leish_023_v3': 625, 'Sample_3-leish_012_v2': 637, 'Sample_10-leish_017_v3': 479, 'Sample_7'：436，'Sample_29-leish032_v2'：659，'Sample_8-leish_016_v3'：375，'Sample_6'：492，'Sample_7-leish_016_v2'：517，'Sample_9'：432，'Sample_8_2'：52 -leish028_v2'：584，'Sample_26-leish027_v3'：629，'Sample_5'：460，'Sample_4'：441，'Sample_3'：668，'Sample_19'：516，'Sample_1'：378，'Sample_2'：437， 'Sample_9-leish_017_v2'：582，'Sample_5-leish_015_v2'：421，'Sample_4-leish_012_v3'：502，'Sample_21-leish_024_v2'：670， 'Sample_18'：573，'Sample_13'：426，'Sample_12'：403，'Sample_11'：463，'Sample_10'：466，'Sample_17'：488，'Sample_16'：487，'Sample_15'：490，'Sample_1 -leish_019_v3'：441，'Sample_30-leish032_v3'：497，'Sample_28-leish028_v3'：542，'Sample_15-leish_021_v2'：837，'Sample_17-leish_022_v2'：747}

Answer 1

您应该能够像这样“解包”字典：

main_key=u'ENSG00000110514'

geo1_labels = data_geo1[main_key].keys() 
geo1_ys = [data_geo1[main_key][x] for x in geo1_labels]
ali_geo1_ys = [data_ali[main_key][x] for x in geo1_labels]

geo2_labels = data_geo2[main_key].keys() 
geo2_ys = [data_geo2[main_key][x] for x in geo2_labels]
ali_geo2_ys = [data_ali[main_key][x] for x in geo2_labels]

然后你可以例如散点图用于比较的值，例如：

import matplotlib.pyplot as plt

fig,ax=plt.subplots()

ax.scatter(range(len(geo1_labels)),geo1_ys,facecolors="None",edgecolors="b",marker="o")
ax.scatter(range(len(geo1_labels)),ali_geo1_ys,facecolors="None",edgecolors="r",marker="s")

ax.set_xticks(range(len(geo1_labels)) )
ax.set_xticklabels(geo1_labels,rotation=90)

plt.tight_layout()
plt.show()

---

更新： 有几种方法可以得到你描述的图，一种方法是使用模块pandas；在这种特定情况下，这似乎需要更多的努力，但它很受欢迎：

import matplotlib.pyplot as plt
import pandas as pd

main_key=u'ENSG00000110514'

compare1=pd.DataFrame.from_dict({
    'data_geo2':data_geo1[main_key],
    'data_ali':{k:data_ali[main_key][k] for k in data_geo1[main_key].keys()}
},orient='index')

compare2=pd.DataFrame.from_dict({
    'data_geo2':data_geo2[main_key],
    'data_ali':{k:data_ali[main_key][k] for k in data_geo2[main_key].keys()}
},orient='index')

compare1['Sample_19-leish_023_v2'].plot.bar()

plt.tight_layout()
plt.show()

另一种选择是直接使用上面的dicts进行绘图：

import matplotlib.pyplot as plt

main_key=u'ENSG00000110514'
sample_key=u'Sample_19-leish_023_v2'

fig,ax=plt.subplots()

xs=[0,1]
ys=[ data_geo1[main_key][sample_key], data_ali[main_key][sample_key] ]

ax.bar(xs,ys)
ax.set_xticks(xs)
ax.set_xticklabels(['data_geo1','data_ali'])

plt.tight_layout()
plt.show()