9.1.1. Components of Floating Point Representation

Now, just look for this particular representation. So, what is the size of this particular representation this is your 32 bits 1 bit is for sign bit, 8 bit is for your exponent and 23 bit is for significant.

Floating point representation is a way to represent real numbers in computers. Each number is encoded using a fixed number of bits. For a standard 32-bit representation, one bit is designated for the sign of the number (indicating if it's positive or negative), eight bits are allocated for the exponent (which determines the scale of the number), and the remaining 23 bits are used for the significand (or mantissa), which contains the actual digits of the number.

Think of floating point representation like a set of address labels used for mail. The sign bit tells you if the mail is an 'addressee' or a 'return to sender' (positive or negative). The exponent helps in scaling where the delivery should be (akin to a postal code), and the significand represents the specific address details (like house number and street name).

So, this is basically same numbers positive and negative positive exponent and negative exponent now what is the exponent over here you just see here I am talking about 2 to the power.

The significand is the part of the floating point number that represents the actual value. In practice, floating point numbers are normalized, which means that the decimal (or binary) point is placed just after the first non-zero digit. For example, a binary number like 1.10100001 is normalized by ensuring it starts with 1 (implicitly), making calculations easier and standardizing the format.

Consider how people often round partial numbers when discussing prices. Instead of saying 1.10100001, someone might say approximately $1.10. Just like rounding makes it easier to communicate about costs, normalization ensures that computer systems can efficiently process floating point numbers by placing the 'decimal point' in a standard position.

Now, just look for the negative now this is a negative exponent this is your -20, but for -20 what we are storing 1 + 2 + 8, 8 + 16 + 32 + 64.

In floating point representation, the exponent can represent both positive and negative values. To handle negative exponents more easily, we use a technique called 'biased exponent.' This means that we adjust the actual exponent value by adding a 'bias' value. For instance, with an 8-bit exponent, the bias is typically 127. Therefore, if we want to represent -20, we would calculate 127 - 20 to find the biased exponent that is stored.

Imagine you have a box that can only show positive temperatures, but you need to read temperatures below zero. Instead of displaying negative values, you could set a reference point—like freezing at 32°F—and start all calculations from there. So, -20°F would be converted to a positive number, making it easier to work with, just as the biased exponent makes handling negative exponents simpler.

Now, what is the accuracy? And accuracy basically talk about what is the changes of the value if we are going to change the least significant bit.

The precision of floating point representation indicates how closely a number can approximate reality. It is determined by the number of bits allocated in the significand. For example, with 23 bits, the accuracy of changes becomes significant; if there's a change in the least significant bit, the value can vary. With floating point numbers, the general understanding is that the more bits you have, the higher the precision in representing values.

Consider a high-resolution camera versus a low-resolution one. A low-resolution camera may capture a picture with visible pixelation or blurriness when zoomed in, while a high-resolution camera maintains sharpness and detail. Similarly, floating point precision determines how accurately we can represent and manipulate numerical values without losing important information.

So, for this particular representation say if I want to store this particular number 1.638125 × 220 basically in binary representation.

The IEEE 754 standard defines how floating point numbers should be represented in computers to ensure consistency and compatibility across different systems. It specifies formats for both 32-bit (single precision) and 64-bit (double precision) numbers which include the layout of the sign bit, exponent, and significand. Utilizing the IEEE 754 standard allows programmers to work with floating point numbers reliably and accurately across various computing platforms.

Think of the IEEE 754 standard as the universal rules for a board game. Just as every player must follow the same rules for a fair and enjoyable game, computers rely on this standard so different systems can interpret and process floating point numbers uniformly, preventing errors and miscalculations.

--