NPTEL Data Analytics with Python Week 1 Assignment Answers | Jan 2024

 

Which of the following is never possible for variance?

 Zero variance

 larger than the standard deviation

 Negative variance

 smaller than the standard deviation

def m(data)

        Diff = max(data) – min(data)

         return(Diff)

The above defined data function in Python programming, will calculate the?

 Inter quartile range

 mode

 median

 range

The function m(data) calculates the range of the data.

Here's a breakdown of how it works:

1. max(data): This part of the code finds the largest value within the dataset.
2. min(data): This part finds the smallest value within the dataset.
3. Diff = max(data) - min(data): This subtracts the smallest value from the largest value, effectively calculating the difference between the two extremes of the data, which is precisely the range.
4. return(Diff): This returns the calculated range value as the output of the function.

So, when you call this function with a set of data, it will provide you with the spread of the data, indicating how far apart the minimum and maximum values are.

Bar Charts are used for :

 Continuous data

 Categorical data

 both a. and b.

 None of these

The correct answer is Categorical data.

Bar charts are best suited for displaying categorical data, meaning data that falls into distinct groups or categories. Each bar represents a different category, and the height or length of the bar corresponds to the frequency or value associated with that category. For example, you could use a bar chart to show the number of students in each grade level at a school or the percentage of each type of animal in a zoo.

While there are technically ways to use bar charts with continuous data (like grouping data into ranges), it can be misleading and difficult to interpret due to the inherent overlap and potential loss of information between data points within each bar. In such cases, visualizations like histograms or line charts are more appropriate.

Therefore, bar charts are primarily used for presenting and comparing categorical data in a clear and visually appealing way.

Frequency polygons are used for: 

 Continuous data

 Categorical data

 both a. and b.

 None of the above

Frequency polygons are primarily used for visualizing continuous data. Here's why:

• Frequency polygons connect the midpoints of class intervals, which only make sense for continuous data divided into specific ranges. Categorical data wouldn't have such intervals.
• Frequency polygons emphasize the overall shape of the distribution, which is more relevant for continuous data where values can theoretically fall anywhere within a range. Categorical data, on the other hand, deals with distinct categories rather than a continuous spectrum.

Here's an example to illustrate:

Continuous data: Imagine you have data on the heights of students in a class. You could group the data into intervals like 150-159 cm, 160-169 cm, etc. A frequency polygon would then connect the midpoints of these intervals (e.g., 154.5 cm for the first interval, 164.5 cm for the second) to show the overall distribution of heights in the class.

Categorical data: Now, suppose you have data on the favorite colors of students in the same class. A frequency polygon wouldn't be a suitable way to visualize this data because colors are not ordered or continuous. Instead, you might use a bar chart to show the number of students who prefer each color.

Therefore, while frequency polygons can technically be used for both continuous and categorical data, they are much more effective and meaningful for visualizing the shape and distribution of continuous data.

I hope this explanation with visuals clarifies the use of frequency polygons!

m is an example of which of the following?

 A population parameter

 sample statistic

 population variance

 mode

Unfortunately, even with the additional information that "m" is a single letter, it's still impossible to determine which of the options is the correct example without any further context.

The terms "population parameter," "sample statistic," "population variance," and "mode" all refer to specific statistical concepts that depend on the data being analyzed. To understand which category "m" falls into, I need to know more about the data and how "m" is calculated.

Here are some questions that might help me understand the context better:

• What kind of data are you working with?
• What is the source of the data? Is it a sample from a larger population, or is it the entire population itself?
• How is "m" calculated? Is it a function of the data, or is it a single value representing something specific?
• What is the purpose of calculating "m"?

Once I have a better understanding of the context, I can provide you with a more accurate answer.

I look forward to hearing more about "m" and helping you solve this problem!

Consider the following statements- 

Statement A : To “flatten” the dataframe, you can use the reset_index(). 
Statement B : Use the nunique() to get counts of unique values on a Pandas Series. 

 Both statements are correct

 Both statements are false

 A is correct, B is false

 B is correct, A is false

The answer is B is correct, A is false.

Here's a breakdown of each statement:

Statement A:

• False. While reset_index() can be used to modify the index of a DataFrame, it doesn't inherently "flatten" the DataFrame in the sense of removing hierarchical structure or converting it to a one-dimensional array.
• To truly flatten a DataFrame, you'd typically use methods like to_numpy() or values.flatten().

Statement B:

• Correct. The nunique() method is specifically designed to count the number of unique values within a Pandas Series.
• It's an efficient way to quickly determine the cardinality (variety) of data within a Series.

Which of the following statements in context of drawing an excellent graph is false:

 The graph should not contain unnecessary adornments (sometimes referred to as chart junk)

 The scale on the vertical axis should begin at zero

 All axes should not be properly labelled

 The graph should contain a title

The statement that is false in the context of drawing an excellent graph is:

The scale on the vertical axis should begin at zero.

While starting the vertical axis at zero is often encouraged for representing data accurately, it can be inappropriate or misleading in certain situations:

• Data with non-zero minimum values: When analyzing data that naturally falls within a specific range (e.g., temperature between 10°C and 40°C), starting the axis at zero can unnecessarily compress the visualization and distort the relative differences between data points.
• Logarithmic scales: For data spanning several orders of magnitude, logarithmic scales are often used to visually compress the vast range but starting at zero wouldn't make sense with log values.
• Emphasis on specific changes: Sometimes, focusing on the relative changes within a specific data range is more important than showing their absolute distance from zero. In such cases, a custom starting point for the vertical axis can highlight the relevant trends more effectively.

Therefore, while avoiding unnecessary clutter and properly labeling axes are crucial for clarity, the decision to start the vertical axis at zero should be based on the specific data and the intended purpose of the graph.

Remember, excellent graphs prioritize informative representation over rigid rules, and adapting to the data is key!

Which of the following is not a measure of dispersion?

 Skewness

 Kurtosis

 Range

 Percentile

Percentile is not a measure of dispersion.

Measures of dispersion quantify how spread out or clustered data points are around the central tendency. While percentiles tell you where specific values fall within a distribution (e.g., the 25th percentile is the value below which 25% of the data points lie), they don't directly measure the spread or variability of the data.

Here's a quick breakdown of why the other options are measures of dispersion:

• Skewness: Shows whether the data is "tilted" to one side of the distribution.
• Kurtosis: Describes the "tail heaviness" of the distribution, indicating how likely extreme values are compared to a normal distribution.
• Range: Simply the difference between the maximum and minimum values, highlighting the overall spread of the data.

Therefore, while percentiles provide valuable information about the distribution, they don't tell you directly how spread out the data points are, making them distinct from measures of dispersion like skewness, kurtosis, and range.

Assume, you are given two lists:

    a = [1,2,3,4,5]

    b = [6,7,8,9]

The task is to create a list which has all the elements of a and b in one dimension.

    Output:

     a = [1,2,3,4,5,6,7,8,9]

Which of the following option would you choose?

 a.append(b)

 a.extend(b)

 Any of the above

 None of these

The correct way to create a list that has all the elements of a and b in one dimension is to use the extend() method as follows:

Python

a.extend(b)

Here's how it works:

1. a.extend(b): This line calls the extend() method on list a and passes b as an argument.
2. extend() function: This function iterates through each element in list b and appends those elements individually to the end of list a.
3. Result: The elements of b are now seamlessly integrated into a, creating a single list with all the elements in the desired order.

After this operation, the contents of list a will be:

a = [1, 2, 3, 4, 5, 6, 7, 8, 9]

Key points to remember:

• extend() directly modifies the original list a, rather than creating a new list.
• append() would have added b as a whole list within a, resulting in a nested structure: [1, 2, 3, 4, 5, [6, 7, 8, 9]].

State the following true or false?

Statement: Bimodal Data sets contains more than two modes

 True

 False

The statement is false. Bimodal data sets contain exactly two modes, not more than two. A mode is the value that occurs most frequently within the data, and a bimodal dataset has two distinct peaks in its distribution, each corresponding to one of the two most frequent values.

Therefore, a dataset can't be bimodal if it has more than two modes. If it contained three or more distinct peaks, it would be considered multimodal, not bimodal.