python vs R

resp=response.dropna(subset=['WorkToolsSelect'])
resp=resp.merge(rates,left_on='CompensationCurrency',right_on='originCountry',how='left')
python=resp[(resp['WorkToolsSelect'].str.contains('Python'))&(~resp['WorkToolsSelect'].str.contains('R'))]
R=resp[(~resp['WorkToolsSelect'].str.contains('Python'))&(resp['WorkToolsSelect'].str.contains('R'))]
both=resp[(resp['WorkToolsSelect'].str.contains('Python'))&(resp['WorkToolsSelect'].str.contains('R'))]
response['LanguageRecommendationSelect'].value_counts()[:2].plot.bar()
plt.show()

重要性

f,ax=plt.subplots(1,2,figsize=(18,8))
response['JobSkillImportancePython'].value_counts().plot.pie(ax=ax[0],autopct='%1.1f%%',explode=[0.1,0,0],shadow=True,colors=['g','lightblue','r'])
ax[0].set_title('Python Necessity')
ax[0].set_ylabel('')
response['JobSkillImportanceR'].value_counts().plot.pie(ax=ax[1],autopct='%1.1f%%',explode=[0,0.1,0],shadow=True,colors=['lightblue','g','r'])
ax[1].set_title('R Necessity')
ax[1].set_ylabel('')
plt.show()

用户比例

from matplotlib_venn import venn2
# pip install matplotlib_venn 画交集用的
f,ax=plt.subplots(1,2,figsize=(18,8))
pd.Series([python.shape[0],R.shape[0],both.shape[0]],index=['Python','R','Both']).plot.bar(ax=ax[0])
ax[0].set_title('Number of Users')
venn2(subsets = (python.shape[0],R.shape[0],both.shape[0]), set_labels = ('Python Users', 'R Users'))
plt.title('Venn Diagram for Users')
plt.show()

py_sal=(pd.to_numeric(python['CompensationAmount'].dropna())*python['exchangeRate']).dropna()
py_sal=py_sal[py_sal<1000000]
R_sal=(pd.to_numeric(R['CompensationAmount'].dropna())*R['exchangeRate']).dropna()
R_sal=R_sal[R_sal<1000000]
both_sal=(pd.to_numeric(both['CompensationAmount'].dropna())*both['exchangeRate']).dropna()
both_sal=both_sal[both_sal<1000000]
trying=pd.DataFrame([py_sal,R_sal,both_sal])
trying=trying.transpose()
trying.columns=['Python','R','Both']
print('Median Salary For Individual using Python:',trying['Python'].median())
print('Median Salary For Individual using R:',trying['R'].median())
print('Median Salary For Individual knowing both languages:',trying['Both'].median())

Median Salary For Individual using Python: 48725.600000000006
Median Salary For Individual using R: 48421.99999999999
Median Salary For Individual knowing both languages: 59791.3

不同工种的偏好

py1=python.copy()
r=R.copy()
py1['WorkToolsSelect']='Python'
r['WorkToolsSelect']='R'
r_vs_py=pd.concat([py1,r])
r_vs_py=r_vs_py.groupby(['CurrentJobTitleSelect','WorkToolsSelect'])['Age'].count().to_frame().reset_index()
r_vs_py.pivot('CurrentJobTitleSelect','WorkToolsSelect','Age').plot.barh(width=0.8)
fig=plt.gcf()
fig.set_size_inches(10,15)
plt.title('Job Title vs Language Used',size=15)
plt.show()

r在视觉上胜过Python。因此，拥有诸如数据分析师、业务分析师等职位头衔的人在图形和视觉上扮演着非常重要的角色，他们喜欢R而不是Python。同样，几乎90%的统计人员使用R，正如前面所述，Python在机器学习方面更好，因此机器学习工程师、数据科学家和DBA或程序员等其他人更喜欢Python。

使用语言多少年啦！

r_vs_py=pd.concat([py1,r])
r_vs_py=r_vs_py.groupby(['Tenure','WorkToolsSelect'])['Age'].count().to_frame().reset_index()
r_vs_py.pivot('Tenure','WorkToolsSelect','Age').plot.barh(width=0.8)
fig=plt.gcf()
fig.set_size_inches(10,10)
plt.title('Job Tenure vs Language Used')
plt.show()

不同的行业用哪门语言

r_vs_py=pd.concat([py1,r])
r_vs_py=r_vs_py.groupby(['EmployerIndustry','WorkToolsSelect'])['Age'].count().to_frame().reset_index()
r_vs_py.pivot('EmployerIndustry','WorkToolsSelect','Age').plot.barh(width=0.8)
fig=plt.gcf()
fig.set_size_inches(10,10)
plt.title('Industry vs Language Used')
plt.show()

r在政府部门还是很强大的，在剩下的其他行业中，Python的份额大约比R高出15-20%。

Python与R分别最常使用的工具是什么呢?

f,ax=plt.subplots(1,2,figsize=(20,15))
py_comp=python['WorkToolsSelect'].str.split(',')
py_comp1=[]
for i in py_comp:
    py_comp1.extend(i)
plt1=pd.Series(py_comp1).value_counts()[1:15].sort_values(ascending=False).to_frame()
sns.barplot(plt1[0],plt1.index,ax=ax[0],palette=sns.color_palette('inferno_r',15))
R_comp=R['WorkToolsSelect'].str.split(',')
R_comp1=[]
for i in R_comp:
    R_comp1.extend(i)
plt1=pd.Series(R_comp1).value_counts()[1:15].sort_values(ascending=False).to_frame()
sns.barplot(plt1[0],plt1.index,ax=ax[1],palette=sns.color_palette('inferno_r',15))
ax[0].set_title('Commonly Used Tools with Python')
ax[1].set_title('Commonly Used Tools with R')
plt.subplots_adjust(wspace=0.8)
plt.show()