Importing data and general info

import pandas as pd
import numpy as np
import re
import matplotlib.pyplot as plt
import seaborn as sns

df = pd.read_csv("data-no-id.csv", encoding='iso-8859-1')

df.columns = ['x', 'y', 'z']

df.head()

	x	y	z
0	0.02	0.00	0.00
1	0.04	0.00	0.20
2	0.06	0.03	0.33
3	0.00	0.03	0.33
4	0.02	0.06	0.00

df.columns

Index(['x', 'y', 'z'], dtype='object')

df['x'].value_counts()

0.02    8
0.06    7
0.04    5
0.51    5
0.00    3
0.13    3
0.09    3
0.17    2
0.11    2
0.28    2
0.53    2
0.55    1
0.62    1
0.36    1
0.15    1
0.91    1
0.87    1
0.72    1
0.19    1
0.21    1
0.98    1
1.00    1
Name: x, dtype: int64

df['y'].value_counts()

0.55    6
0.03    6
0.61    6
0.06    4
0.10    3
0.13    3
0.00    3
0.52    3
0.19    3
0.35    2
0.39    2
0.45    2
0.16    2
0.94    1
1.03    1
0.65    1
0.74    1
0.77    1
1.10    1
0.58    1
0.48    1
Name: y, dtype: int64

df['z'].value_counts()

0.13    9
0.07    8
0.00    6
0.40    5
1.00    5
0.73    4
0.93    2
0.20    2
0.33    2
0.67    2
0.47    2
0.60    2
0.87    1
0.53    1
0.27    1
0.80    1
Name: z, dtype: int64

2d plots of all axes separately

x axis

import seaborn as sns

sns.distplot(df['x'], kde_kws={'clip': (0.0, 1.0)})

/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2557: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).
  warnings.warn(msg, FutureWarning)

<AxesSubplot:xlabel='x', ylabel='Density'>

import seaborn as sns
sns.set_palette("dark")

sns.histplot(df['x'])

<AxesSubplot:xlabel='x', ylabel='Count'>

y axis

import seaborn as sns

sns.distplot(df['y'], kde_kws={'clip': (0.0, 1.0)})

/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2557: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).
  warnings.warn(msg, FutureWarning)

<AxesSubplot:xlabel='y', ylabel='Density'>

import seaborn as sns

sns.histplot(df['y'])

<AxesSubplot:xlabel='y', ylabel='Count'>

z axis

import seaborn as sns

sns.distplot(df['z'], kde_kws={'clip': (0.0, 1.0)})

/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2557: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).
  warnings.warn(msg, FutureWarning)

<AxesSubplot:xlabel='z', ylabel='Density'>

import seaborn as sns

sns.histplot(df['z'])

<AxesSubplot:xlabel='z', ylabel='Count'>

3d plots of all axes

fig = plt.figure()
ax = fig.add_subplot(projection='3d')

x, y, z = np.loadtxt('data-no-id.csv', delimiter=',', unpack=True)

ax.scatter(x,y,z)
plt.show()

Exploration Stage

The main goal here is to explore the next relationships:

• X versus Y
• Y versus Z
• X, Y, and Z

#let's plot it
sns.distplot(df['x'], kde_kws={'clip': (0.0, 1.0)}, color="seagreen")
sns.distplot(df['y'], kde_kws={'clip': (0.0, 1.0)}, color="purple")

/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2557: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).
  warnings.warn(msg, FutureWarning)
/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2557: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).
  warnings.warn(msg, FutureWarning)

<AxesSubplot:xlabel='y', ylabel='Density'>

Comparing the 'X' and the 'Y', we gather the following insights:

• Both X and Y have a similar progressions, as in when X is higher, Y is also high, and vice versa. For our case, this linear relationship means that whenever others ask us about our gender pronouns, we are also more proactive and ask people their pronouns.
• On the other hand, the trend on the 'X' parameter is showing that the distribution is highly skewed to the right side, meaning that majority of people are not asked (0-0.2) about their gender pronoun.
• At the same time, the inversed case for 'Y', is that majority of people either don't ask (0-0.2) about gender pronoun, or they OFTEN (0.4-0.6) ask.

sns.distplot(df['y'], kde_kws={'clip': (0.0, 1.0)}, color="seagreen")
sns.distplot(df['z'], kde_kws={'clip': (0.0, 1.0)}, color="purple")

/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2557: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).
  warnings.warn(msg, FutureWarning)
/usr/local/lib/python3.9/site-packages/seaborn/distributions.py:2557: FutureWarning: `distplot` is a deprecated function and will be removed in a future version. Please adapt your code to use either `displot` (a figure-level function with similar flexibility) or `histplot` (an axes-level function for histograms).
  warnings.warn(msg, FutureWarning)

<AxesSubplot:xlabel='z', ylabel='Density'>

In case of comparing Y (how often do we ask gender pronouns) and Z(our general awareness about gender pronouns), we gathered the following insights:

• People who are less aware of gender pronouns, obviously, are not asking others about their gender pronouns.
• An unexpected insight, is that there are more people people who are OFTEN (0.4-0.6) asking about gender pronouns, then those who consider themselves aware about gender pronouns. In our opinion, this insight could be contributed to either 1)a misunderstanding in our data collection method or 2)subjective evaluation of what it means to be aware of gender profiles.
• finally, we can observed higher density of people who consider themselves highly aware of gender pronouns, yet are not used to asking others about their gender pronoun, which indicates a non-proactive behavior.

The idea is to map the X variable (how much we ask others about their gender pronoun) against the other 2.

Our hypotheses are:

• X variable would be influence by the Y variable (if others ask us, there is a higher probability that we will ask others)
• Z variable (if we are more aware abou gender profiles), there is more probability that we'll ask others.

plt.style.use("ggplot")
fig, axs = plt.subplots(2,2, figsize = (25,15))
fig.suptitle("How often do we ask about gender profiles", fontsize = 20)

sns.regplot(y = 'y', x = 'x', data = df, ax = axs[0, 0])
sns.regplot(y = 'y', x = 'z', data = df, ax = axs[0, 1])
axs[1,2].set_xticklabels(axs[1,2].get_xticklabels(), rotation = 45, horizontalalignment = 'right')

plt.show()

---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
<ipython-input-125-9c8ebf331a15> in <module>
      5 sns.regplot(y = 'y', x = 'x', data = df, ax = axs[0, 0])
      6 sns.regplot(y = 'y', x = 'z', data = df, ax = axs[0, 1])
----> 7 axs[1,2].set_xticklabels(axs[1,2].get_xticklabels(), rotation = 45, horizontalalignment = 'right')
      8 
      9 plt.show()

IndexError: index 2 is out of bounds for axis 1 with size 2

From the 2 graphs we can conclude the following information:

• the relationship between X and Y is indeed linear, and we are asking others more about their pronouns if they asked us first
• on the other hand, we found that the general awareness and our proactivity to ask do not have a linear relationship and the plot is highly scattered. If anything, there is a negative behavior -- if we are more aware about gender pronouns, we ask people less than if we are unaware.

correlation matrix

corr = df.corr()
mask = np.zeros_like(corr, dtype=np.bool)
mask[np.triu_indices_from(mask)] = True
cmap = sns.diverging_palette(225, 7, as_cmap=True)

fig, axs = plt.subplots(1,1, figsize = (15,15))
sns.heatmap(corr, mask = mask, cmap = cmap)

<ipython-input-126-8497c6b7606c>:2: DeprecationWarning: `np.bool` is a deprecated alias for the builtin `bool`. To silence this warning, use `bool` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.bool_` here.
Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations
  mask = np.zeros_like(corr, dtype=np.bool)

<AxesSubplot:>

Finally, to confirm all of the above, we've plotted the correlation matrix between all of the varibale and found:

• X and Y indeed show high correlation.
• whilst Z and Y show are skewed toward a negative correlation, which we've already discussed in previous points.