Consulting - Quality
Difference makes the DIFFERENCE
Agriculture DataSet
Bars, Boxes and Functions
In [1]:
!pip3 install plotly_express
In [2]:
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline
import plotly_express as px
In [4]:
df = pd.read_csv('/content/clean_agri_ds.csv')
In [5]:
df.State_Name.unique()
Out[5]:
In [6]:
df.dtypes
Out[6]:
In [7]:
df.head(3)
Out[7]:
In [8]:
df[df.isna().any(axis=1)]
Out[8]:
In [9]:
df[df['Production'] == "="]
Out[9]:
In [10]:
df.drop(['Len State', 'Len Dist', 'Len Season', 'Len Crop'], inplace = True, axis=1)
In [11]:
df.head(3)
Out[11]:
In [12]:
df.describe()
Out[12]:
In [13]:
df.set_index(['Unnamed: 0'])
Out[13]:
In [14]:
df.rename(columns = {'Unnamed: 0':"ColIndex"}, inplace = True)
In [15]:
df.head(2)
Out[15]:
In [16]:
df.set_index(['ColIndex'])
Out[16]:
In [17]:
df.head(2)
Out[17]:
In [18]:
df.sort_values(['Production'], inplace = True, ascending = False)
In [19]:
df.head(2)
Out[19]:
Create subset for AP from India dataset¶
In [20]:
dfap = df[df['State_Name'] == "Andhra Pradesh"]
In [21]:
dfap.head(3)
Out[21]:
In [22]:
dfap_group = df[df['State_Name'] == "Andhra Pradesh"].groupby('District_Name').sum()
In [23]:
dfap_group
Out[23]:
In [24]:
dfap.head(2)
Out[24]:
In [25]:
dfap.rename(columns = {"Unnamed: 0":"ColIndex"}, inplace=True)
In [26]:
dfap.set_index('ColIndex')
Out[26]:
In [27]:
dfap.describe()
Out[27]:
In [28]:
dfap['Crop_Year'] = df['Crop_Year'].astype(str)
In [29]:
dfap['ColIndex'] = dfap['ColIndex'].astype(str)
In [30]:
dfap.describe()
Out[30]:
In [31]:
sns.lineplot(x="District_Name", y="Area", hue='Crop_Year', data = dfap)
plt.xticks(rotation = 90)
plt.legend(bbox_to_anchor = [1,1])
Out[31]:
In [32]:
sns.lineplot(x="District_Name", y="Production", hue='Crop_Year', data = dfap)
plt.xticks(rotation = 90)
plt.legend(bbox_to_anchor = [1,1])
Out[32]:
In [33]:
sns.barplot(x="District_Name", y="Area", data = dfap)
plt.xticks(rotation = 90)
Out[33]:
In [34]:
sns.barplot(x="District_Name", y="Production", data = dfap)
plt.xticks(rotation = 90)
Out[34]:
In [35]:
dfap['District_Name'].unique()
Out[35]:
In [36]:
dfap.describe()
Out[36]:
In [37]:
dfap.dtypes
Out[37]:
Working with Plot features in Pandas¶
- The .plot.* methods are applicable on both Series and DataFrames
- By default, each of the columns is plotted as a different element (line, boxplot,…)
- Any plot created by pandas is a Matplotlib object.
source: https://pandas.pydata.org/docs/getting_started/intro_tutorials/04_plotting.html
In [38]:
dfap.plot(figsize = (8,4))
Out[38]:
In [39]:
dfap.groupby('District_Name')['Area'].sum().plot.bar()
Out[39]:
In [40]:
dfap.groupby('District_Name').sum().plot(kind='line')
plt.xticks(rotation=90)
Out[40]:
In [41]:
dfap.sort_values(by="District_Name", inplace=True)
In [42]:
dfap.plot.scatter(x='Area', y='Production')
Out[42]:
In [43]:
dfap.plot.box()
Out[43]:
In [44]:
fig, axs = plt.subplots(figsize = (12, 6));
# dfap.plot.area(figsize = (12, 6), subplots = True)
dfap.plot.line(ax=axs, subplots = True, marker=".");
In [45]:
fig.savefig('Area_Production.png')
In [46]:
dfap
Out[46]:
In [47]:
dfap.sort_values("Production", ascending = True, inplace = True)
In [48]:
dfaps = dfap['Production'].cumsum()
In [49]:
dfaps.plot()
plt.xticks(rotation = 90)
Out[49]:
In [50]:
dfapa = dfap['Area'].cumsum()
In [51]:
dfapa.plot()
plt.xticks(rotation = 90)
Out[51]:
In [52]:
plt.figure()
dfap.Production.plot(kind='density')
Out[52]:
In [53]:
df2 = pd.DataFrame(np.random.rand(10, 4), columns = ['a', 'b', 'c', 'd'])
df2.plot.bar();
In [54]:
df2 = pd.DataFrame(np.random.rand(10, 4), columns = ['a', 'b', 'c', 'd'])
df2.plot.bar(stacked = True);
In [55]:
df2 = pd.DataFrame(np.random.rand(10, 4), columns = ['a', 'b', 'c', 'd'])
df2.plot.bar(stacked = True);
plt.legend(bbox_to_anchor = [1,1]) # to place legend outside the graph, towards the right top
Out[55]:
In [56]:
df2 = pd.DataFrame(np.random.rand(10, 4), columns = ['a', 'b', 'c', 'd'])
df2.plot.barh(stacked = True);
In [57]:
dfap['Production'].diff().hist()
Out[57]:
In [58]:
plt.figure()
dfap['Area'].diff().hist(color = 'k', bins=50)
Out[58]:
In [59]:
plt.figure()
dfap['Area'].diff().hist(color = 'k', bins=50)
Out[59]:
In [60]:
dfapb = dfap.sample(1000).groupby('Season').sum()
dfapb.plot.box()
Out[60]:
In [61]:
dfapb
Out[61]:
In [62]:
dfapb = dfap.sample(10)
dfapb.boxplot()
Out[62]:
In [63]:
dfap.sample(10).boxplot(by='Crop_Year', figsize=(10, 3))
plt.xticks(rotation=90)
Out[63]:
In [64]:
dfap.sample(10).boxplot(by='Season', figsize=(10, 3))
plt.xticks(rotation=90)
Out[64]:
In [65]:
dfap.sample(10).boxplot(by='Crop', figsize=(10, 3))
plt.xticks(rotation=90)
Out[65]:
In [66]:
dfap.head(2)
Out[66]:
In [67]:
dfap.sample(10).boxplot(by="Season");
plt.xticks(rotation=90);
In [68]:
dfap.sample(10).boxplot(by=["Season", 'Crop'],
figsize = (8,6), layout=(2, 1));
plt.xticks(rotation=90);
In [69]:
dfap.plot.scatter(x="Production", y='Area')
Out[69]:
In [70]:
dfap.plot.scatter(x="Production", y='Area', s=5)
Out[70]:
In [71]:
<