anygraph/app.py at main · SamuelHLewis/anygraph · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
import streamlit as st
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

def load_user_data(filename):
    if filename is not None:
        return pd.read_csv(filename)
    return None

# make sure to keep this as the first call, otherwise the plot will be tiny
st.set_page_config(layout='wide')

# this will be used to infer the correct chart type
continuous_types = ['int32', 'int64', 'float32', 'float64']

# define the available criteria as a session state
# this prevents new criteria from overwriting existing ones
if 'criteria_options' not in st.session_state:
    st.session_state.criteria_options = []

# define the total number of criteria as a session state
# this allows criteria levels to be labelled with unique and meaningful keys
if 'total_criteria_levels' not in st.session_state:
    st.session_state.total_criteria_levels = 1

# we want the plot (middle column) to be thrice as large as the menus
# we deliberately place the menus outside of the sidebar to separate tabs from page-specific elements
col1, col2 = st.columns([1,2])

# lefthand column holds the plotting options
with col1:
    user_file = st.file_uploader("Choose a file", type=['csv'])
    input_df = load_user_data(user_file)

    if input_df is not None:
        x_var = st.selectbox(
            "X Axis",
            input_df.columns,
            index=None,
            placeholder="Select a variable",
            key='x_selector'
        )
        y_var = st.selectbox(
            "Y Axis",
            input_df.columns,
            index=None,
            placeholder="Select a variable",
            key='y_selector'
        )
        # this section creates a third box to customise the plot in a plot-type-specific manner
        if x_var == None and y_var == None:
            pass
        # for scatterplots, offer a colour selector
        elif x_var and y_var and input_df[x_var].dtype in continuous_types and input_df[y_var].dtype in continuous_types:
            colour_var = st.selectbox(
                "Point Colour",
                input_df.columns.drop([x_var, y_var]),
                index=None,
                placeholder="Select a variable",
                key='colour_selector'
            )
        # for histograms, offer a granularity (i.e. total bins) slider
        elif x_var == None or y_var == None:
            total_bins = st.slider(
                "Granularity",
                min_value=2,
                max_value=int(len(input_df)),
                value=int(len(input_df)/5)
            )
        # for continuous vs categorical, offer a radio button to toggle between boxplot and violinplot
        elif x_var and y_var and (
            (input_df[x_var].dtype in continuous_types and input_df[y_var].dtype not in continuous_types) or
            (input_df[x_var].dtype not in continuous_types and input_df[y_var].dtype in continuous_types)):
            cont_vs_cat_plot_type = st.radio(
                "Plot type",
                ["Box-Whisker Graph", "Violin Graph"]
            )

    # righthand column holds the graph itself
    if input_df is not None:
        with col2:
            fig, ax = plt.subplots()
            if x_var == None and y_var == None:
                st.write("Select which variables you would like to visualise using the dropdowns on the left")
            elif y_var == None:
                ax = sns.histplot(data=input_df, x=x_var, bins=total_bins)
            elif x_var == None:
                ax = sns.histplot(data=input_df, y=y_var, bins=total_bins)
            else:
                x_type = input_df[x_var].dtype
                y_type = input_df[y_var].dtype
                if x_type in continuous_types and y_type in continuous_types:
                    ax = sns.scatterplot(data=input_df, x=x_var, y=y_var, hue=colour_var)
                elif (x_type in continuous_types and y_type not in continuous_types) or (
                    x_type not in continuous_types and y_type in continuous_types):
                    if cont_vs_cat_plot_type == "Box-Whisker Graph":
                        ax = sns.boxplot(data=input_df, x=x_var, y=y_var)
                    elif cont_vs_cat_plot_type == "Violin Graph":
                        ax = sns.violinplot(data=input_df, x=x_var, y=y_var)
                elif x_type not in continuous_types and y_type not in continuous_types:
                    total_counts = input_df[[x_var, y_var]].value_counts().reset_index()
                    total_counts_matrix = total_counts.pivot(index=x_var, columns=y_var, values="count")
                    ax = sns.heatmap(total_counts_matrix)
            st.pyplot(fig)