科技3D视界 · Bilibili · 9:18
This 120-second video intuitively explains area and volume using 'uniformly distributed points' and 'points swept by moving shapes'. First, in two dimensions, translating a line segment yields rectangle area; rotating a line segment about one endpoint yields circle area, and the midpoint trajectory converts circle area into rectangle area πr². Then it generalizes to three dimensions: translating a square yields rectangular prism volume; rotating a square about one side yields cylinder volume, and the same midpoint trajectory idea gives πr³. Finally, the video cuts the square along its diagonal into right triangles, rotating one into a cone and translating the other into a prism, raising the analogy 'cone volume = prism volume ???'; this statement is not proven within the segment and contradicts standard formulas. The video opens with a logical paradox concerning the inconsistency between the volume formulas of cones and prisms. It then formally introduces matrices, vividly demonstrating through animated grid transformations their power as tools for spatial transformation. The video compares two-dimensional rotation matrices, Euler's formula, and rotation matrices in three and four dimensions, highlighting the universality and advantages of matrices in expressing rotations in higher-dimensional spaces. Finally, it shows how matrices can represent the Minkowski spacetime light cone, concluding with suspense to prompt viewers to reflect on the essence of matrices. This segment explains the origin of matrices using a system of three linear equations: first, each equation is viewed as the dot product of a coefficient vector and the unknown vector; then, the three coefficient vectors are stacked to form the coefficient matrix, yielding the matrix-times-vector form. The video then summarizes matrix multiplication as representing linear transformations and illustrates its intuitive effect with animations of a two-dimensional grid—namely, rotation and stretching while preserving uniform spacing of unit intervals along axes. Finally, it proposes replacing the coordinate axes with and , leaving open the question of how such axis changes differ from linear transformations. This segment first constructs a two-dimensional "sine coordinate system" using and , reading the point [2, 3] as ; then adds to obtain a three-dimensional example [1, 1, 1]→. Next, the video generalizes this interpretation to infinite dimensions, writing , and names it the sine form of the Fourier series. Finally, the video adopts the phrasing "viewing continuous functions directly as vectors", states that all continuous functions constitute a complete Euclidean space, and introduces the term "Hilbert space", ending with biographical background information. This segment mainly introduces the famous 'Hilbert 23 Problems' in the history of mathematics. The video first highlights the strong influence of this school by displaying portraits of Hilbert and his students—Weyl, von Neumann, and Wiener. Then, a scrolling list presents the specific content and solution status of these 23 problems; the narration especially mentions the Continuum Hypothesis and the Riemann Hypothesis. The video discusses the great significance of compiling such a list, pointing out that it requires the proposer to possess deep expertise across all branches of mathematics. Finally, the video reviews the earlier material on matrices and Hilbert spaces, and concludes with a thought-provoking question about the volume of a solid of revolution as homework.
Use the learning inspector for key ideas and moments, or open the reading tabs for the complete notes.
Generated from the video's visuals and explanation; not verbatim speech.
At the beginning, space is intuitively set up as "flat space", and an animation shows many white dots transitioning from random scattering to a regular grid, illustrating that points are considered uniformly distributed in this model. This step establishes the premise for later interpreting area and volume as "the number of points swept".
Next, the scene shifts to a two-dimensional plane. A yellow line segment appears on the grid and moves upward, sweeping out a rectangle filled with diagonal lines. The video thus interprets area as the points swept by the moving line segment, so rectangular area equals "length of the segment times distance moved".
The question then becomes: what if this line segment rotates one full turn around one of its endpoints instead of translating? The animation first draws the resulting circle, then points out that the number of points swept differs at different radii. The video handles this by taking the trajectory of the segment's midpoint as the "average distance moved", and writes on screen: "midpoint trajectory = ".
Once this average distance is established, the video treats the circle equivalently as a rectangle with base r and height πr, and directly writes "circle area = πr²" and "rectangle area = πr²". The reasoning order here is: first acknowledge that point-sweeping under rotation is non-uniform, then reinterpret it as comparable translational area using the midpoint trajectory.
Then the video extends the same idea into three dimensions. A cyan square translates perpendicularly within coordinates, sweeping out a rectangular prism. Volume is thus defined as the total number of points swept by the moving square, expressed as "square area times distance moved".
Next, the square rotates one full turn around one of its sides, generating a cylinder. The video explicitly states "according to our earlier logic", again applying the midpoint trajectory formula "", treating the cylinder as a rectangular prism with square base and length πr, thus obtaining "cylinder volume = rectangular prism volume = πr³".
Finally, the video cuts the square along its diagonal into two right triangles, rotating the left one around one of its legs to form a cone, and translating the right one to form a triangular prism. The voiceover emphasizes: although the cross-sectional area is now only half the original square's, at every instant the area of the rotating right triangle remains exactly half the square's area, so by analogy "the number of points swept through space should also be equal".
The clip ends with the on-screen text "Volume of a cone = Volume of a prism ???". This should be understood as an interrogative analogy posed by the video, not a proven theorem; to assess its validity, one must refer back to the standard volume formulas for independent verification.
At the beginning of the video, an aqua wireframe cone and a triangular prism appear on screen. The narrator presents a seemingly rigorous logical argument: if the reasoning is flawless, the volume of the cone should equal that of the prism. Then the screen displays their respective volume formulas: and —clearly unequal. The video leaves this contradiction as a 'homework problem', prompting viewers to reflect on whether spatial rotation affects volume.
The scene shifts to a two-dimensional grid. The narrator formally introduces the core concept of 'matrix'. Through a series of animations, it demonstrates how different matrices stretch, shear, and rotate the entire grid space. The narrator emphasizes that matrices are a powerful tool for studying spatial transformations—and even mentions Einstein’s use of them in deriving general relativity, establishing matrices’ pivotal role in advanced mathematics and physics.
To illustrate concretely, the video presents the two-dimensional rotation matrix , accompanied by an animation of a line segment rotating. At this point, the narrator raises a common question: Euler’s formula can also represent rotation. Yet the video immediately points out that Euler’s formula becomes unwieldy when representing rotations of vectors in three-dimensional space—thus highlighting the necessity of matrices.
The video then showcases the three-dimensional rotation matrix . In the animation, a line segment rotates about the z-axis, generating the cone shown at the beginning—tying back to the earlier setup. Next, matrices are extended to four dimensions , illustrating that even though humans cannot intuitively visualize four-dimensional space, matrices still precisely describe rotations within it.
Finally, the video applies matrices to physics, displaying the matrix representation of the Minkowski spacetime light cone. The image renders a symmetric double cone, top and bottom. The narrator quotes the famous line from *The Three-Body Problem*: 'Within the light cone lies destiny', linking mathematical tools to cosmic philosophy. At the end, the screen dissolves into falling digital rain, leaving a lingering question about the true nature of matrices.
At the beginning of the segment, digital rain and a continuously changing array of numbers create a visual impression of 'matrix', but the voiceover immediately clarifies that matrices are not mysterious: they originally arose from solving systems of multivariate equations. Here, the digital rain serves only as an introduction and does not perform any specific computation.
Next, the screen displays a system of three linear equations in three unknowns: , , . The video first examines the first equation alone, identifying as the dot product of the coefficient vector and the unknown vector , with the right-hand side being the scalar result of this dot product.
Then, the three equations are viewed as three dot products: the coefficient vector of each row takes the dot product with the same column vector of unknowns, yielding . Visually, these results are stacked vertically into a column vector, while the three rows of coefficients are stacked vertically into the matrix, so the entire system naturally rewrites as .
On this matrix equation, the video names three objects: the array on the left is the "coefficient matrix", the unknown column vector in the middle is the "pre-transformation vector", and the resulting column vector on the right is the "post-transformation vector". Thus, matrix multiplication is interpreted as transforming one vector into another, a process called a "linear transformation" or "linear mapping".
The scene then cuts to a split-screen style reminiscent of The Matrix, using the title 'Matrix' and numeric matrices as an analogy. The narration points out that the film's title 'Matrix' literally means 'matrix', and the relationships in the film's world can be understood as a kind of projection; however, linear transformations in mathematics are not as magical as in the film—they are described here primarily as rotations and stretches.
To visualize these abstract ideas, the segment transitions to an animated two-dimensional blue grid. In the upper-left corner, matrices , , , , and so on—each an matrix—appear sequentially, causing the grid to skew, rotate, or stretch. A key invariant is that unit markings along the axes remain evenly spaced: the distance from 1 to 2 equals the distance from 2 to 3.
The segment ends with a challenge: what happens if we move beyond linear grid deformations and instead let the coordinate axes themselves become functions such as and ? The screen displays "x-axis: " and "y-axis: ". This segment concludes here, presenting only this idea as a lead-in to the next level of distinction, without giving a full answer.
At the beginning of the segment, the ordinary Cartesian coordinate system is transformed into a 'sine coordinate system': the horizontal axis is no longer the ordinary x-direction but ; the vertical axis is no longer the ordinary y-direction but . Consequently, the grid in the image changes from straight lines to wavy ones; the focus is not on plotting function graphs, but on letting the coordinate axes themselves be determined by functions.
Next, the video selects a sample point within this two-dimensional sine coordinate system. Here, the term 'point' does not denote a position in the plane, but rather the coefficients in front of the two sine axes, so it is read as . The image simultaneously displays a 'combined waveform' to illustrate that this expression corresponds to a superimposed function.
Then, the author extends the same idea to three dimensions by adding a axis . Now the three coordinate axes are , , and , and the spatial grid in the image becomes denser and more distorted, illustrating how intuition declines as dimension increases.
In the three-dimensional example, the video selects the point and reads it, following the same rule, as . This step is entirely isomorphic to the earlier two-dimensional example: the coordinate components remain the coefficients of the sine 'basis axes'.
Next, the video extends 'finite-dimensional coordinate points' further to 'infinite-dimensional coordinate vectors', displaying a column vector with ellipsis on screen and writing . The author directly calls this infinite sum 'the sine form of the Fourier series'. The key here is formal analogy: infinitely many coordinate components correspond to infinitely many sine terms.
Based on this identification, the video further infers: if Fourier series can be analyzed in this way, then "almost all continuous functions" can also undergo similar analysis, because Fourier series can synthesize other functions. This passage is the author's verbal generalization; the video does not present a convergence theorem or a rigorous proof.
Then the video rephrases the idea: rather than drawing skewed coordinate axes, one should directly regard any continuous function itself as a vector. In this view, the object of discussion shifts from "points in a curved coordinate system" to "elements in a function space".
Under this new perspective, the video claims that all continuous functions together form a "complete Euclidean space", and names this space the "Hilbert space". The words "Hilbert space" appear prominently on screen. This is a conceptual introduction, not a rigorous definition; inner products, norms, and completeness are not elaborated in this segment.
Finally, the clip shifts to historical background, showing portraits of Hilbert, Hermann Weyl, and John von Neumann to illustrate the historical origin of the term "Hilbert space". No new mathematical formulas or derivations appear here.
The video opens with portraits of four renowned mathematicians: Hilbert, Weyl, von Neumann, and Wiener. The narration points out that in a certain important ranking—contextually inferred to be the ranking of the greatest mathematicians of the 20th century—Hilbert and his students occupy four of the top ten positions, vividly illustrating the profound influence of the Hilbert school in mathematics.
Next, the screen shifts to a scrolling yellow-on-black list, displaying the specific contents of "Hilbert's 23 Problems" and their current resolution status (e.g., "partially solved", "solved", "unsolved"). The narration explains that at the International Congress of Mathematicians in 1900, Hilbert posed these 23 most important mathematical problems. The list specially highlights the first problem, "Cantor's Continuum Hypothesis", and the eighth problem, "the Riemann Hypothesis".
Addressing the question of whether merely listing problems is significant, the narration gives an affirmative answer. Because mathematics comprises numerous branches and contains an overwhelming number of problems, selecting the most important ones demands that the proposer possess comprehensive mastery of all mathematical branches and their historical proofs. Such broad academic vision and deep expertise are beyond the reach of ordinary mathematicians—this is precisely what makes Hilbert's act so remarkable.
At the end of the video, the narration summarizes this episode's brief introduction to "matrices" and "Hilbert spaces", and previews that the next episode will delve further into matrices. It also leaves an intriguing homework question: Why is the volume formed by rotating half of a square not equal to half the volume of a rectangular prism moved the same distance? The video then concludes with a farewell remark.
The video first intuitively sets space as "flat space", and uses an animation showing points evolving from random scattering to a regular grid to illustrate that, in this model, all points are uniformly distributed. This assumption is later used to support the claim that "area/volume equals the number of points swept by a moving figure".
In the two-dimensional plane, the video interprets area as the set of points swept by a moving line segment. If the segment translates upward, the region swept is a rectangle; thus, area equals "length of segment × translation distance".
When a line segment rotates about one endpoint, points at different radii sweep different numbers of points. The video instead uses the circular trajectory length of the segment's midpoint as the average translation distance, writing "midpoint trajectory = ", then treats the circle as a rectangle with base r and height πr, yielding circle area πr².
Upon entering three dimensions, the video interprets volume as the set of points swept by a moving square. Translating the square perpendicularly forms a rectangular prism, so volume equals "area of square × translation distance".
Rotating a square about one of its edges generates a cylinder. Following the two-dimensional approach, the video takes the midpoint trajectory "" as the average translation distance, thus expressing cylinder volume as the volume of a rectangular prism with square base and length πr, i.e., πr³.
At the end of the video, a square is cut into right triangles: one rotates about a leg to form a cone, and another translates to form a prism; then, by analogy with the fact that "the area of a triangle is always half that of the square," the question "volume of cone = volume of prism ???" is raised. This is merely a speculative inference at the end of the segment; the video does not provide a proof.
The video twice uses "taking the midpoint as the average" to convert rotation-sweep problems into translation problems, but never proves why this averaging is valid. When studying, treat it as an intuitive analogy rather than a complete derivation.
The equation "volume of cone = volume of prism ???" appears on screen with a question mark, indicating that the video only poses the question. According to standard formulas, if both legs are r, the cone's volume is ()πr³, while the triangular prism's volume is r³; these are generally unequal.
At the start of the video, a logical inference is proposed: the volume of a cone should equal that of a prism. Yet the given formulas are and —unequal. This contradiction serves as a hook to spark reflection on spatial transformations.
A matrix is not merely a table of numbers—it is a mathematical tool for describing spatial deformations such as stretching, rotating, and shearing. Through grid animations, the video vividly shows how different matrices reshape two-dimensional space, and also references their application in general relativity.
Two-dimensional rotation can be represented either by matrix or by Euler’s formula . However, the video notes that Euler’s formula fails when handling vector rotations in three or higher dimensions, whereas matrices extend effortlessly.
The video demonstrates three-dimensional and four-dimensional rotation matrices. A three-dimensional matrix rotates a line segment about an axis to generate a cone; although four-dimensional matrices are difficult for humans to visualize, mathematically they simply add another dimension and still accurately describe rotations in higher-dimensional space, illustrating the powerful generality of matrices.
A matrix involving hyperbolic functions and can represent the light cone in Minkowski spacetime in relativity. Combining dialogue from The Three-Body Problem, the video points out that the light cone separates the fate perceivable by humans from the unknowable dark universe, showcasing the application of matrices in modeling physical spacetime.
The video first places matrices back into their historical motivation: matrices are not mysterious symbols appearing out of nowhere, but rather tools developed to handle systems of multivariable equations concisely. The array in the opening digital rain is merely a visual hook; the actual mathematical content begins with a system of three linear equations in three variables.
Taking as an example, the left-hand side can be written as . The coefficients form a row vector, the unknowns form a column vector, and the two vectors have matching dimensions; their dot product equals the constant on the right-hand side.
Each of the three equations corresponds to a dot product. Arranging the three rows of coefficients into a matrix and the three results into a column vector yields . Each row of the matrix dotted with the column vector of unknowns recovers the original equation.
The video summarizes matrix multiplication as a mapping: it takes a vector as input and produces another vector as output. Here, the "pre-transformation vector" is the column vector of unknowns, and the "post-transformation vector" is the resulting column vector; this mapping is called a linear transformation or linear mapping.
The English title of The Matrix shares the word 'matrix' with the mathematical concept; this video uses that coincidence to aid memory, but immediately clarifies that linear transformations in mathematics serve a more down-to-earth purpose: in grid animations, they mainly produce rotation and stretching—not the surreal spatial powers seen in the film.
A two-dimensional grid becomes skewed or stretched under the action of several matrices, yet unit intervals along any single coordinate axis remain equally spaced. The video uses 'the distance from 1 to 2 equals the distance from 2 to 3' as an intuitive criterion for linearity.
At the end of this segment, the idea is raised of using as the horizontal axis and as the vertical axis. This contrasts with the earlier linear grid property of preserving uniform spacing along axes—but this segment only poses the question, without delivering a full conclusion.
The video sets the horizontal axis as and the vertical axis as , causing the entire grid to bend accordingly. Here the emphasis is not on plotting functions within a fixed coordinate system, but on using the functions themselves as coordinate directions.
In this two-dimensional sine coordinate system, is read as . That is, the coordinate components represent not planar positions, but weights along each sine axis.
After adding as the -axis, the three coordinate axes become , , and . The spatial grid shown on screen becomes more complex, illustrating the increased intuitive burden upon raising the dimension.
Point is read as following the same rule. This is a direct extension of the reading method used in the two-dimensional example.
Explore conditions, steps and evidence. Supplementary explanations are labeled separately from content shown in the video.
The screen marks the radius r on a two-dimensional diagram showing a circle and a rectangle, and writes below it "the midpoint's path = 2 times r divided by 2 times pi equals pi r."
r
The length of the rotating line segment in the two-dimensional example, which is the radius of the resulting circle
positive real number
Next to a three-dimensional diagram showing a cylinder and a rectangular prism, the screen writes "the midpoint's path = 2 times r divided by 2 times pi equals pi r," and labels the rectangular prism's "distance moved" as pi r.
r
In the three-dimensional example, when a square rotates around one of its sides, the square's side length also serves as the radius of rotation
positive real number
The screen displays "Volume of a cone = ".
The cone volume expression given in the video.
Volume of a geometric solid.
The screen displays "Volume of a prism = ".
The prism volume expression given in the video.
Volume of a geometric solid.
The screen displays the matrix [Cos(θ), -Sin(θ); Sin(θ), Cos(θ)].
[Cos(θ), -Sin(θ); Sin(θ), Cos(θ)]
Rotation matrix in two-dimensional space.
Linear algebra / geometric transformation.
The screen displays , where θ varies with the animation.
Euler's formula, used to represent rotation in the two-dimensional plane.
Complex numbers / geometric transformation.
The screen displays an matrix [Cos(θ), -Sin(θ), 0; Sin(θ), Cos(θ), 0; 0, 0, 1]
[Cos(θ), -Sin(θ), 0; Sin(θ), Cos(θ), 0; 0, 0, 1]
Matrix for rotation about the z-axis in three-dimensional space
Linear algebra / geometric transformations
The screen displays an matrix [Cos(θ), -Sin(θ), 0, 0; Sin(θ), Cos(θ), 0, 0; 0, 0, 1, 0; 0, 0, 0, 1]
[Cos(θ), -Sin(θ), 0, 0; Sin(θ), Cos(θ), 0, 0; 0, 0, 1, 0; 0, 0, 0, 1]
Rotation matrix in four-dimensional space
Linear algebra / higher-dimensional geometry
The screen displays an matrix [Cosh(θ), -Sinh(θ), 0, 0; Sinh(θ), Cosh(θ), 0, 0; 0, 0, 1, 0; 0, 0, 0, 1]
[Cosh(θ), -Sinh(θ), 0, 0; Sinh(θ), Cosh(θ), 0, 0; 0, 0, 1, 0; 0, 0, 0, 1]
Matrix used to represent the light cone in Minkowski spacetime
Relativity / hyperbolic functions
The screen displays a system of three linear equations in three variables: , , .
An example of a multivariable linear system used to introduce matrices
x, y, and z are the unknowns
The first equation is rewritten as [3 2 1]·[x y z]^.
[3\ 2\ 1]
the coefficient vector of the first equation
a three-dimensional row vector
The unknowns x, y, and z are written vertically as a column vector.
the column vector of unknowns
a three-dimensional column vector
Voiceover: "Notice that, in the general case, we always regard the space we inhabit as a flat space—that is, in this space all points are uniformly distributed."
On a black background, numerous small white dots spread outward from the center, then rearrange into a regular grid pattern.
The video first intuitively sets space as "flat space" and explains this assumption using the phrase "all points are uniformly distributed." This statement is used in this segment to support later interpretations of area and volume as the "number of points swept."
The video uses it as an intuitive premise in the general case
no rigorous mathematical definition provided
The narrator says: "Now, let's look at a two-dimensional plane... So when we compute area, we can think of it as the set of points swept out by a line segment as it moves."
A yellow horizontal line segment appears in a two-dimensional grid and, after moving upward, forms a rectangle filled with diagonal lines.
The screen labels "length of the segment" and "distance moved".
The video interprets two-dimensional area as the total number of points swept out by a line segment moving within the plane. In the case of translation, the swept region is a rectangle, so the area is written as "length of the segment × distance moved".
Applies to the intuitive two-dimensional plane model shown in the video
A line segment forms a rectangle by translating perpendicularly.
The narrator says: "What if this line segment rotates once around one of its endpoints? Although the curved path means points at different distances from the center sweep different numbers of points, if we take the midpoint of AB as an average, the points swept by this line segment remain uniformly distributed; thus, the area of this circle equals the area of a rectangle whose translation distance is the radius's midpoint."
A yellow line segment rotates once around its left endpoint to form a circle; simultaneously, a rectangle appears on the right with height r and width πr.
The screen displays "midpoint's trajectory = ", and also writes "circle area = πr²" and "rectangle area = πr²".
The video treats "taking the midpoint of AB as the average" directly as justification that the swept points remain uniform, without providing a rigorous proof.
The video first acknowledges that when a line segment rotates about one endpoint, the number of points swept differs across different radii; it then adopts the circular trajectory length πr of the segment's midpoint as the 'average moving distance', equating the disk area to the area of a rectangle with base r and height πr, thereby obtaining πr².
The line segment has length r.
The line segment rotates one full turn about one endpoint.
The video uses the midpoint's trajectory as an intuitive assumption for the average moving distance.
The narrator says: "Okay, let's go one step deeper into three-dimensional space. When we translate a square by some distance, we obtain a rectangular prism. When computing volume, we can imagine calculating the total number of points swept by the square during its motion. So the volume equals the square's area multiplied by the distance it moves."
In three-dimensional coordinates, a cyan square appears and, after translating vertically, forms a rectangular prism.
The diagram labels "square side length" and "translation distance".
The video extends the two-dimensional idea "a line segment sweeps points to yield area" to three dimensions: the total number of points swept by a translating square equals the volume, so the volume equals the square's area times the translation distance.
Applies to the three-dimensional intuitive model shown in the video.
The square translates perpendicularly to its own plane to form a rectangular prism.
The narrator says: "Similarly, if we rotate this square around one of its sides for one full turn, then according to our earlier reasoning, its volume should equal the volume of a rectangular prism whose base is the square and whose height is the circular circumference traced by the square's midpoint radius."
The cyan square rotates one full turn about one of its vertical sides to form a cylinder; on the right appears a rectangular prism whose base is a square and whose length is labeled πr.
The screen displays: "midpoint trajectory = ", "distance traveled = πr", "cylinder volume = rectangular prism volume = πr³".
The video adopts the two-dimensional idea of "midpoint trajectory as average distance traveled", but does not prove that sweeping points in three-dimensional rotation can be averaged in the same way.
The video analogizes the cylinder formed by rotating the square about one of its sides to a rectangular prism with the square as its base and the circular trajectory of the midpoint, πr, as its "distance traveled", thus obtaining volume ² = πr³.
The square has side length r.
The square rotates one full turn about one of its sides.
The video uses the intuitive assumption that the midpoint trajectory represents the average distance traveled.
The voiceover says: "Then what if we cut this square in half? Like this—we then obtain a cone and a triangular prism. Although the square's area is halved, at every instant during rotation the area of the right triangle being rotated is always half that of the square, so the points swept out in space—whether rotating in a circle or moving straight—should also be equal."
Two cyan squares are each divided by a diagonal into right triangles; the left triangle rotates about one of its legs to form a cone, while the right triangle translates perpendicularly to form a triangular prism.
The screen finally displays: "cone volume = triangular prism volume ???".
The video presents this conclusion with a question mark at the end of the segment, without providing proof or numerical verification.
This conclusion contradicts the standard formula: if both legs are r, then the volume of the cone is ()πr³, whereas the volume of the triangular prism is r³.
The video extends the earlier 'sweeping point' logic: since the area of the right triangle used for rotation remains always half that of the original square, the number of spatial points swept should be equal whether rotating in a circle or translating in a straight line; thus it raises the question-like conclusion 'volume of cone = volume of prism'.
Right triangle obtained by cutting a square along its diagonal, serving as the generating cross-section
Left side: rotation about one leg; right side: translation perpendicular to the plane
The video only raises the question, without proof.
The narrator says, "Matrices are a powerful tool for studying spatial transformations."
Animation showing a grid being stretched, skewed, and rotated by different matrices
The video uses animation to show how different matrices transform the shape of a two-dimensional grid, pointing out that matrices are a powerful tool for studying spatial transformations and mentioning that Einstein used this tool to derive general relativity.
Suitable for describing spatial transformations
The screen displays an rotation matrix
A line segment in the coordinate system rotates as the matrix parameters change
The video gives the formula for a rotation matrix in two-dimensional space and uses animation to demonstrate how this matrix rotates a line segment in a two-dimensional coordinate system.
Acts on a two-dimensional vector
The screen displays an rotation matrix
A line segment in a three-dimensional coordinate system rotates about the z-axis, sweeping out a conical surface
The video shows a rotation matrix in three-dimensional space and uses animation to demonstrate how this matrix rotates a line segment about the z-axis, thereby generating a cone.
Acts on a three-dimensional vector
The screen displays the rotation matrix
The narrator says, "Matrices can rotate things just as well."
The video presents the rotation matrix formula in four-dimensional space, explaining that even though humans cannot intuitively visualize four-dimensional space, matrices can still represent rotational transformations within it.
Acts on a four-dimensional vector
The screen displays an matrix containing hyperbolic functions
Generates a vertically symmetric double cone (light cone) in a three-dimensional coordinate system
The video demonstrates using a matrix containing hyperbolic cosine and hyperbolic sine to represent the light cone in Minkowski spacetime, accompanied by a quotation from The Three-Body Problem about fate inside and outside the light cone and the dark universe.
Used for relativistic spacetime geometry
The voiceover says, "Matrices were originally developed to solve systems of multivariable equations."
The background features digital rain, and the foreground displays a dynamically changing numeric matrix.
This video places matrices back into their original motivation: handling systems of multivariable equations in a compact way. The numeric array shown on screen serves only as a visual introduction; the actual mathematical object is then introduced via concrete systems of equations.
The subject under discussion is systems of multivariable linear equations.
The voiceover says: 'So the area of this circle equals the area of a rectangle whose translation distance is the radius's midpoint.'
The screen displays 'circle area = πr²' and 'rectangle area = πr²'.
In the intuitive model of this video, the area of a circle with radius r equals the area of a rectangle with base r and height πr, i.e., πr².
Segment length is r
Segment rotates one full turn about one endpoint
using the midpoint trajectory 〔πr〕 as the average moving distance
holds for the circle of radius 〔r〕 shown in the video
The narrator says: "Its volume should equal that of a rectangular prism whose base is a square and whose moving distance is the circumference traced by the midpoint of the radius."
The screen displays "cylinder volume = rectangular prism volume = 〔πr³〕."
In the intuitive model of this video, the volume of the cylinder generated by rotating a square of side length 〔r〕 about one of its sides equals the volume of a rectangular prism with that square as base and length 〔πr〕, namely 〔πr³〕.
the square has side length 〔r〕
the square rotates one full turn about one of its sides
using the midpoint trajectory 〔πr〕 as the average moving distance
holds for the solid of revolution of the square of side length 〔r〕 shown in the video
The narrator says: "Therefore, whether it moves in a circular path or translates straight, the set of points swept out in space should be equal."
The screen displays "cone volume = prism volume ???".
The video poses the question in the form of a question mark, without providing a proof.
According to standard geometric formulas, this equation generally does not hold.
At the end of the video, a question is raised: given two right triangles cut from the same square, one rotated about a leg to form a cone and the other translated perpendicularly to form a prism, are their volumes equal?
The right triangle is obtained by cutting the square along its diagonal.
The left figure rotates about one leg.
The right figure translates perpendicularly.
The video does not explicitly quantify the range; it raises the question only through an illustrative diagram.
The narrator states, "If human logic hasn’t gone wildly astray, then the volume of the cone should equal the volume of the prism."
The screen contrasts the cone volume with the prism volume
The video does not provide specific dimensional parameters for the prism, only displaying the formula result.
Based on a certain logical derivation, the volume of the cone should equal that of the prism, but the actual computed formula results are inconsistent.
Human logic has not committed an egregious error.
Universal quantifier
The voiceover explicitly states, "Matrices were originally developed to solve systems of multiple equations."
Matrices were originally developed to solve systems of multiple equations.
The video context introduces the origin of matrices
Regarding the origin of matrices as described in the video
The voiceover says a single equation can be viewed as the dot product of a coefficient vector and an unknown vector.
The visual rewrites as the dot product of a row vector and a column vector.
A linear equation can be viewed as the dot product of a coefficient vector and an unknown vector equaling the right-hand constant.
The equation is a linear equation.
Coefficients and unknowns are arranged in the same order.
This holds for each equation in the example.
The voiceover says that the three equations naturally represent three inner products, whose results, when arranged together, form a three-dimensional vector.
The right-hand side 39, 34, 26 are boxed as a column vector.
The three equations of a system of three linear equations correspond respectively to three inner products, and the three results, when arranged together, form a three-dimensional vector.
The system contains three linear equations.
The unknowns are x, y, and z.
This holds for the example system of equations.
The voiceover concludes, "Ultimately, a matrix serves to transform one vector into another vector."
Ultimately, a matrix serves to transform one vector into another vector.
Matrix-vector multiplication exists.
The video presents this in a general statement.
The narrator says that the effect of a linear transformation on space is merely rotation and stretching.
The grid undergoes shearing, rotation, and stretching.
"Merely rotation and stretching" is an intuitive summary from the video, without strict constraints or a complete classification given in this segment.
The video states that the effect of a linear transformation on space is merely rotation and stretching.
Within the context of two-dimensional grid visualization.
Regarding the linear transformation shown in the animation.
The narrator imagines using for the horizontal axis and for the vertical axis, thereby questioning the change in coordinate axes.
The screen displays "horizontal axis: " and "vertical axis: ".
The video does not fully explain why this coordinate axis substitution is not a linear transformation before the segment ends.
The video proposes using for the horizontal axis and for the vertical axis as an idea for how "coordinate axes also keep changing".
Coordinate axes parameterized by nonlinear functions
A counterexample idea that viewers might propose
The narrator says, "It is precisely the sine form of the Fourier series."
The screen simultaneously displays .
The video does not give the standard definition of the sine form of the Fourier series.
It does not explain the origin of the coefficients a or the convergence conditions of the series.
The video claims that the expression , obtained by extending finite-dimensional sine coordinates to infinite dimensions, is the "sine form of the Fourier series."
We have accepted the video's earlier interpretation of coordinate components as coefficients with respect to sine basis functions.
The space dimension is extended to infinite dimensions.
Regarding the infinite-dimensional sine combination form shown in the video,
The narrator says, "If the Fourier series can be analyzed in this way, then it effectively means that almost all continuous functions admit such analysis, because the Fourier series can synthesize other functions."
"Almost all" is not quantitatively defined in the video.
The video does not specify which convergence theorem or function-class conditions are required.
This statement is an inference drawn in the video, not a proof given in the video.
The video infers: if Fourier series can be analytically analyzed in this manner, then "almost all continuous functions" can be analyzed this way, because Fourier series can synthesize other functions.
Fourier series can be analytically analyzed in the manner shown in the video.
Fourier series can synthesize other functions.
The video uses the phrase "almost all continuous functions" but gives no rigorous quantifier or description of the exceptional set.
The voiceover explains step by step: during rotation, points at different radii trace different numbers of points; the midpoint is taken as the average value; then the area of the circle is converted into the area of a rectangle.
The screen gradually displays: "midpoint trajectory = ", "area of circle = πr²", "area of rectangle = πr²".
A line segment rotates into a circle, while an equal-area rectangle appears on the right.
Taking the midpoint as the average value is merely an intuitive assumption made in the video, with no rigorous derivation provided.
First consider a line segment of length r rotating about one endpoint, sweeping out a circle.
The video screen and voiceover directly present this generation process.
The video acknowledges that the rotation path is a curve, and the number of points swept varies at different distances from the center.
The voiceover explicitly states this difficulty.
Instead, use the circumference length traced by the segment’s midpoint as the 'average moving distance'.
The on-screen formula gives this directly.
Treat the circle equivalently as a rectangle with base r and height πr, yielding area πr².
The screen simultaneously displays 'circle area = πr²' and 'rectangle area = πr²'.
The video intuitively derives circle area as πr² using the midpoint trajectory.
The voiceover extends the two-dimensional reasoning to three dimensions, explaining that the volume generated by rotating a square about one side equals the volume of the corresponding rectangular prism.
The screen displays 'midpoint trajectory length = ', 'moving distance = πr', 'cylinder volume = rectangular prism volume = πr³'.
A square rotates to form a cylinder, and a rectangular prism of equal volume appears synchronously on the right.
This derivation adopts the two-dimensional 'midpoint averaging' assumption, but the video does not prove that the same assumption holds in three dimensions.
A square with side length 〔r〕 rotates about one of its sides to generate a cylinder.
The video's visuals and narration present this directly.
The trajectory length of the midpoint along the square's width direction is taken as the average distance traveled.
The formula appears directly on screen.
The cylinder is treated equivalently as a rectangular prism with base area 〔r²〕 and length 〔πr〕, yielding volume 〔πr³〕.
The screen displays 'cylinder volume = rectangular prism volume = 〔πr³〕'.
Using the same midpoint trajectory idea as for circle area, the video intuitively derives cylinder volume as 〔πr³〕.
The narration states that after cutting the square in half, at any moment during rotation, the area of the right triangle being rotated is always half the area of the original square, so the points swept through in space should also be equal in measure.
Two right triangles undergo rotation and translation respectively, generating a cone and a triangular prism.
The screen ends with "Volume of a cone = Volume of a prism ???".
The video presents only an analogy argument and ends with a question mark, without completing the proof.
This conclusion conflicts with the standard volume formula; check later content to see whether it is corrected.
First cut the original square cross-section in half to obtain a right triangle.
The video screen directly shows the slicing process.
The video emphasizes that at any moment during rotation, the area of the rotating right triangle is always half that of the square.
The voiceover explicitly states this proportional relationship.
By analogy, the number of spatial points swept out by circular rotation should equal that swept out by linear translation.
The voiceover gives the analogical conclusion, and the screen displays it with a question mark.
At the end of the segment, the video raises the analogical question "Volume of a cone = Volume of a prism", but does not prove it.
The screen sequentially displays the volume formulas for the cone and the prism for comparison.
List the volume formula for the cone.
shown directly in the video
List the volume formula for the prism.
shown directly in the video
The results of the two volume formulas are inconsistent ( versus ), raising a logical question.
Rewriting the system of equations row by row into inner products, then combining them into matrix multiplication.
The narration proceeds in the order: "single equation—three inner products—coefficient matrix—result vector".
Starting from a three-variable linear system.
The example system of equations given directly in the video.
Viewing the first-row equation as the inner product of a coefficient row vector and the unknown column vector.
Using the definition of the vector inner product.
The three equations respectively give three inner product results, arranged vertically into a column vector.
The video arranges the three right-hand constants together to form a three-dimensional vector.
Arrange the three coefficient vectors row-wise into a coefficient matrix, thereby writing the entire system of equations in matrix multiplication form.
Each row of the matrix multiplied by the vector is exactly the inner product of the corresponding equation.
A three-variable linear system can be equivalently written as the coefficient matrix times the unknown column vector equals the result column vector.
The blue grid becomes a slanted, rotated, or stretched grid under the action of several matrices.
The voiceover emphasizes that unit scaling remains uniformly distributed.
The grid tilts under the action of a shear-type matrix.
The video demonstrates the effect of a linear transformation using the matrix in the upper-left corner together with grid deformation.
The grid undergoes an overall turning resembling rotation.
The video continues demonstrating rotation and stretching using different matrices.
No matter how the grid is skewed or stretched, the distance between adjacent unit markings on the same axis remains constant.
The voiceover explicitly states that the distance from 1 to 2 and the distance from 2 to 3 remain equal.
The video uses a grid animation to illustrate that linear transformations can change direction and scale, but do not make unit intervals along coordinate axes non-uniform.
The screen sequentially displays [2; 3], 2Sin(x)+3Sin(2x), [1; 1; 1], Sin(x)+Sin(2x)+, [1; 1; 1; ...], aSin().
The voiceover progresses from two dimensions to three dimensions to infinite dimensions, and finally states, 'It is the sine form of the Fourier series.'
The generalization from finite terms to infinite terms lacks rigorous justification.
The convergence of the series and the method for determining coefficients are not explained.
First, in the two-dimensional sine coordinate system, the point [2; 3] is interpreted as a linear combination of coefficients along two sine axes.
The coordinate reading directly given in the video
Then extend the same reading to three dimensions by adding the axis , yielding a three-term sine combination.
The three-dimensional extension directly given in the video
Finally raise the dimension further to infinite dimensions, representing infinitely many components with an ellipsis vector and writing the corresponding infinite sum.
The infinite-dimensional formal extension directly given in the video
The video names this infinite sum "the sine form of the Fourier series".
The spoken conclusion in the video matches the on-screen formula.
Through a continuous analogy—"two-dimensional coordinate point → three-dimensional coordinate point → infinite-dimensional coordinate vector"—the video extends finite-term sine combinations to infinite-term sine series and identifies them as the sine form of the Fourier series.
The voiceover first criticizes "such a wobbly coordinate system is just too ugly", then states "we can regard any continuous function directly as a vector… all continuous functions together form a complete Euclidean space… called Hilbert space".
The visual shifts from a complex twisted wireframe grid back to a relatively straight grid, and the words "Hilbert space" appear.
This derivation is a conceptual restatement, not a rigorous proof.
No definition is given for the inner product or completeness of the function space.
The video first points out that the earlier approach of bending coordinate axes into sine curves is overly complicated and insufficiently intuitive.
The narrator explicitly comments, "A wobbly coordinate system is simply too ugly."
Instead, the viewpoint shifts: rather than emphasizing deformed coordinate axes, continuous functions themselves are treated directly as vectors.
The narrator says, "We can regard any continuous function directly as a vector."
Under this viewpoint, all continuous functions are described as forming a complete Euclidean space.
The narrator states this conclusion directly.
Finally, the video names this space the "Hilbert space."
The screen displays the large text "Hilbert space," while the narrator simultaneously explains the origin of the name.
The video replaces the earlier intuitive drawing with bent coordinate axes by adopting the approach of "regarding functions themselves as vectors," thereby introducing the term "Hilbert space."
The voiceover first discusses uniform point distribution in a two-dimensional plane, then area via line-segment translation, and finally area of a circle via line-segment rotation about one endpoint.
The yellow line segment is translated across the grid to form a rectangle, then rotated about one endpoint to form a circle, and the circle is placed side by side with the rectangle for comparison.
On screen: 'segment length', 'translation distance', 'midpoint trajectory length = ', 'circle area = πr²', 'rectangle area = πr²'.
In the two-dimensional plane, how can we understand rectangle area and circle area using the concept of 'swept points'?
Points in the plane are assumed to be uniformly distributed.
There is a line segment of length r.
The segment can be translated upward or rotated fully around one endpoint.
This yields expressions for rectangle area and circle area, respectively.
Translating the segment sweeps out a rectangle whose area equals the segment length times the translation distance.
The video narration and on-screen labels state this directly.
When rotating the segment about one endpoint, the length of the midpoint's trajectory is taken as the average translation distance.
The on-screen formula states this directly.
Treat the circle as a rectangle with base r and height πr to obtain the circle area.
The screen displays "Circle area = πr²".
Rectangle area is "segment length × distance moved"; circle area is πr².
The video self-checks by writing πr² for both circle area and rectangle area side by side.
The narration extends the two-dimensional logic into three dimensions: first, a square translates to form a rectangular prism; then, a square rotates about one of its sides to form a cylinder.
A cyan square translates into a rectangular prism in three-dimensional coordinates, then rotates about one of its sides to form a cylinder, and is placed side by side with the rectangular prism for comparison.
The screen shows "Square side length", "Distance moved", "Midpoint trajectory = ", "Cylinder volume = Rectangular prism volume = πr³".
In three-dimensional space, how can we understand rectangular prism volume and cylinder volume using the concept of "swept points"?
Points in space are assumed to be uniformly distributed.
There is a square with side length r.
The square can translate perpendicularly or rotate fully about one of its sides.
We obtain the expressions for the volume of a rectangular prism and the volume of a cylinder, respectively.
Translating a square sweeps out a rectangular prism; its volume equals the area of the square multiplied by the distance moved.
The video narration and on-screen labels give this directly.
When rotating a square about one of its sides, we take the length of the trajectory of its midpoint as the average distance moved.
The on-screen formula gives this directly.
We treat the cylinder as equivalent to a rectangular prism with base area 〔r²〕 and length 〔πr〕, yielding its volume.
The screen displays "volume of cylinder = volume of rectangular prism = 〔πr³〕".
The volume of the rectangular prism is "area of square × distance moved"; the volume of the cylinder is 〔πr³〕.
The video performs a self-check by writing the cylinder volume and the rectangular prism volume side by side as 〔πr³〕.
The narration asks what happens if we cut the square in half, and states that this yields a cone and a triangular prism.
Two right triangles, rotated about and translated along one of their legs, generate a cone and a triangular prism, respectively.
The screen finally shows "volume of cone = volume of prism ???".
This example does not give the final answer within the segment; it only poses the question.
If a square is cut along its diagonal into two right triangles, and one right triangle rotates about one of its legs while the other translates perpendicularly, are their resulting volumes equal?
The square is cut along its diagonal into two right triangles.
The left right triangle rotates one full turn about one of its legs.
The right right triangle translates perpendicularly.
Compare the volumes of the resulting cone and prism.
The video emphasizes that at every moment during rotation, the area of the rotating right triangle equals half the area of the square.
The voice-over states this explicitly.
By analogy, the number of points swept out by rotation and by translation should be equal.
The voice-over gives an analogy, and the screen displays a question mark.
The segment does not give a definite answer; it only raises the question "volume of cone = volume of prism ???".
None; the video intentionally leaves a question mark here without verification.
two dimensional grid under different 2times2 matrices undergoes skewing, stretching, and rotation
observe the effect of different matrices on the two dimensional space grid
initial standard two dimensional grid
Multiple distinct matrices
Intuitively understand the role of matrices as tools for spatial transformations.
Demonstrate a shear transformation matrix.
Video footage
Demonstrate a rotation/reflection transformation matrix.
Video footage
Matrices can significantly alter the shape of space and are powerful tools for studying spatial transformations.
Animations visually demonstrate changes in grid shape.
The screen fully displays a system of three linear equations in three variables and its matrix form.
The voiceover explains the origin of the matrix with respect to this system of equations.
Given the system of equations , , , how do we write it in matrix form?
The system of equations is a linear system.
The unknowns are x, y, and z in that order.
The right-hand side constants are 39, 34, and 26 in that order.
Rewrite the system as the coefficient matrix multiplied by the column vector of unknowns.
First consider the first equation and recognize its left-hand side as the dot product of the coefficient vector and the unknown vector.
Definition of the inner product.
Arrange the coefficients of the three equations row-wise to obtain the coefficient matrix.
The video states that the three coefficient vectors placed together form a third-order matrix.
Place the unknowns column vector to the right of the matrix, and arrange the three inner product results into a result column vector.
The row-wise inner products of matrix multiplication with a vector correspond to the original system of equations.
A three-by-three matrix, with first row three, two, one; second row two, three, one; third row one, two, three; multiplied by the column vector consisting of x, y, z equals the column vector consisting of thirty-nine, thirty-four, twenty-six.
The inner product of each row of the matrix with [x, y, z]^T recovers , , and , matching the original system of equations.
The screen shows [2; 3] and 2Sin(x)+3Sin(2x)
The narrator says, "For example, the point [2, 3] corresponds to the vector ."
The screen simultaneously displays "combined waveform" and an overlaid oscillating curve.
The exact image parameters of this combined waveform are not provided.
It is not explained how to infer coordinates from the waveform.
In the two-dimensional sine coordinate system with horizontal axis and vertical axis , take the point [2, 3] and find its corresponding "vector" expression.
The coordinate axes are defined as horizontal axis and vertical axis .
The example point is .
Write this point as its corresponding sine combination expression.
First read off the two components of this point in the two-dimensional sine coordinate system.
The example point directly given in the video.
Use the first component 2 as the coefficient of and the second component 3 as the coefficient of to obtain the combined expression.
The video's reading rule for this coordinate system.
2 times sine x plus 3 times sine 2x
The video hints at the connection between this expression and sine series using a 'combined waveform' diagram and the voiceover 'Could this be... a Fourier series?', but does not provide numerical verification.
The screen displays [1; 1; 1] and
The voiceover says, 'Let’s take just three 1s; then the vector it represents is '
The exact parameters of these three waveforms are not given.
It is not explained why [1; 1; 1] is chosen as the example.
In a three-dimensional sine coordinate system with axes , , and , take the point [1, 1, 1] and find its corresponding 'vector' expression.
The three coordinate axes are , , and .
The example point is
Write this point as the corresponding three-term sine combination expression.
First read off the three components of this point in the three-dimensional sine coordinate system.
example points given directly in the video
Since all three components are 1, simply add the three sine axes to obtain the three-term combination.
the video's rule for reading this coordinate system
sine x plus sine 2 x plus sine 3 x
The video then extends this reading rule to infinite dimensions and uses it to support the claim 'Fourier series in sine form', but does not separately verify this three-term expression.
White dots spread outward from the center on a black background, then form a regular grid.
Voiceover explains "flat space" and "all points are uniformly distributed."
White scattered dots.
Regular lattice grid.
Dots spread outward from the center.
The arrangement of dots changes from loose to a regular grid.
The background remains black.
The dots remain white.
Use uniform arrangement of dots to intuitively represent uniform distribution of points in "flat space".
A yellow horizontal line segment appears in the two-dimensional grid, and after upward translation, forms a rectangle filled with diagonal lines.
On-screen labels: "segment length" and "distance moved".
Yellow line segment.
Rectangle filled with diagonal stripes.
Grid background.
Line segment translated upward.
Swept region gradually fills to form a rectangle.
Segment remains horizontal during translation.
Grid remains unchanged.
Interpret area as the total number of points swept by the segment moving in the plane.
Yellow line segment rotates one full turn about its left endpoint to form a circle; simultaneously, a rectangle appears on the right with height r and width πr.
On-screen text: "midpoint trajectory = ", "circle area = πr²", "rectangle area = πr²".
Rotating the line segment
Circle
Rectangle
Radius labeled r
Midpoint trajectory formula
Line segment rotating about one endpoint
Circle gradually forming
Rectangle on the right appears simultaneously with the formula
Length of the line segment remains r
Rotation center remains fixed
Using the midpoint trajectory πr to convert the circle's area into the area of an equal-area rectangle.
In three-dimensional coordinates, a cyan square appears and, after translating vertically, forms a rectangular prism.
The screen labels "side length of the square" and "distance moved".
cyan square
rectangular prism
three dimensional coordinate axes
square translated in the vertical direction
the swept region gradually becomes a rectangular prism
the square's shape and size remain unchanged
the coordinate axes remain unchanged
Interpret volume as the total number of points swept by the square moving through space.
The cyan square rotates one full turn about one of its vertical edges to form a cylinder; on the right, a rectangular prism appears with a square base and length labeled 〔πr〕.
On-screen text: "Midpoint trajectory = ", "Travel distance = πr", "Cylinder volume = Rectangular prism volume = πr³".
Rotating the square.
Cylinder.
Rectangular prism.
Midpoint trajectory formula.
Square rotating about one of its sides.
Cylinder forming gradually.
Rectangular prism on the right appears simultaneously with the formula.
Side length of the square remains r.
Axis of rotation remains fixed.
Use the midpoint trajectory πr to convert the cylinder volume into the volume of an equal-volume rectangular prism.
Two cyan right triangles undergo rotation and translation respectively: the left one forms a cone, and the right one forms a triangular prism.
The screen finally displays "Volume of cone = Volume of prism ???".
Two right triangles
Cone
Triangular prism
Question-mark conclusion text
Left triangle rotates about one of its legs
Right triangle translates perpendicularly
Both solids gradually take shape
Both triangles are cut from the same square
The area of each triangle remains half that of the square
The video uses an analogy of halving area to raise the question of whether the volumes of a cone and a prism are equal.
The left is a cyan wireframe cone; the right is a cyan wireframe triangular prism.
Cone.
Triangular prism.
No dynamic change; static display.
The geometric shapes of the two solids.
Raising the question of whether their volumes are equal.
A line segment rotating in a two-dimensional coordinate system as the angle increases, with 〔〕 displayed beside it.
Coordinate axes.
Line segment.
Euler's formula text.
The line segment's angle increases over time
The θ value in changes in real time
The line segment's length remains constant
Demonstrating the correspondence between two-dimensional rotation and Euler's formula
In a three-dimensional coordinate system, the line segment rotates about the z-axis, and its trajectory forms a dashed conical surface
Three-dimensional coordinate axes
Rotating line segment
Dashed conical trajectory
The line segment continuously rotates about the z-axis
The angle between the line segment and the z-axis remains unchanged
Intuitively demonstrating how a three-dimensional rotation matrix generates a cone
generate a cyan wireframe double cone symmetric about the horizontal and vertical planes in three dimensional coordinate system with a yellow plane in the middle
Double cone (light cone)
Yellow horizontal plane
three dimensional coordinate axes
The light cone gradually unfolds and takes shape
The light cone's vertex lies at the origin
Illustrating the spacetime light cone structure in relativity
On a dark background, green falling digits form numeric matrix suspended at the center, with values continuously changing.
These random digits carry no explicit mathematical meaning and serve primarily as a visual cue.
Green falling digits.
The central numeric matrix.
The numbers in the matrix are continuously updated
The matrix always maintains an arrangement
Introduce the topic using the visual symbol of a 'matrix', then transition to a concrete system of equations.
Each equation in the system is highlighted row by row, with coefficients, unknowns, and right-hand constants separately boxed.
A three-variable system of equations
Yellow coefficient box
Cyan unknown box
Green result box
First highlight the coefficients and unknowns of a single equation
Then group the three rows of coefficients, the three columns of unknowns, and the right-hand results into vectors or matrices respectively
The equations themselves remain unchanged
Keep the unknowns in the order x, y, z.
Visually decompose the algebraic equation into a coefficient vector, an unknown vector, and a result vector, preparing for matrix notation.
The video twice uses 'taking the midpoint as the average' to convert a rotation-sweeping-point problem into a translation problem.
write directly on the screen "midpoint movement trajectory = 2 times r divided by 2 times pi equals pi r", and use this to derive the area of a circle and the volume of a cylinder.
Seeing the video directly obtain πr² and πr³ from the midpoint trajectory, viewers may mistakenly think this is already a rigorous proof.
This segment only provides an intuitive analogy: it first acknowledges that points swept by different radii differ during rotation, then directly uses the midpoint trajectory length as the 'average moving distance'. The video does not prove that such averaging is valid in both two and three dimensions.
The voiceover says, 'The points swept out in space should also be equal.'
The screen ends with "Volume of a cone = Volume of a prism ???".
It is easy to misinterpret the question-mark equation at the end of the segment as a conclusion already proven by the video.
Here the video only raises an analogy and a question, without giving a proof; moreover, according to standard formulas, if both legs of the right triangle are r, then the cone's volume is ()πr³ while the triangular prism's volume is r³, and these two are generally unequal.
The voiceover says, "But what if we have a vector in three-dimensional space? Then Euler's formula is not well suited for expressing it."
The belief that Euler's formula can perfectly express spatial rotations in any number of dimensions.
The video points out that Euler's formula has limitations when expressing vector rotations in three or more dimensions, whereas matrices can easily extend to any number of dimensions.
the narrator says, "There is nothing mysterious happening here."
matrices appear to be a mysterious or seemingly arbitrary symbolic system.
the video emphasizes that matrices originally arose from systems of multiple equations; once we return to their origin in such systems, the matrix notation feels natural.
the narrator says the effect of a linear transformation on space is not as magical as in movies.
because the movie is titled 'The Matrix', people mistakenly assume linear transformations possess the magical spatial abilities seen in the film.
the video uses the movie as a naming analogy while pointing out that, mathematically, linear transformations on a grid mainly manifest as rotations and stretches, preserving uniform unit spacing along the axes.
The narrator simulates an audience member saying, "I insist on letting the coordinate axes themselves vary and change," and gives examples such as and .
This segment ends here, without providing a complete rebuttal or subsequent definition.
One could arbitrarily replace the coordinate axes with functions such as or , and still regard this as producing the same effect as the linear transformations described earlier.
The video uses this idea to prompt distinguishing linear transformations from more general coordinate-axis replacements; this segment only raises the question and does not develop a full conclusion.
the video says "all continuous functions form a complete Euclidean space … called a Hilbert space"
The screen only gives the name "Hilbert space", without definitions of inner product, norm, or completeness
need to check the original video to see whether a rigorous definition is provided elsewhere
After hearing 'the set of all continuous functions forms a complete Euclidean space', one might mistakenly think that any collection of continuous functions is naturally a Hilbert space.
The analyst adds: a Hilbert space usually requires first specifying an inner product, inducing a norm from it, and then requiring the space to be complete under that norm. The video here gives only an intuitive statement, without providing these structures, so it cannot be taken as a rigorous definition.
The video says 'it is the sine form of the Fourier series' and 'almost all continuous functions admit such an analysis'.
The screen displays only aSin(), with no convergence conditions or coefficient formulas.
It is unknown whether the original video later supplements a convergence theorem.
Upon seeing the extension from finite sine combinations to infinite series, one might mistakenly think the video has already proved that every continuous function can be expanded into a Fourier series.
The analyst adds: the video performs only formal analogy and naming, without giving the Fourier coefficient formulas, a convergence theorem, or a rigorous meaning of 'almost all', so this is not a proof.
The voiceover says 'for example, the point [2, 3], then the vector it expresses is2sinx+3sin2x'.
The screen displays column vectors alongside function expressions.
In this segment, [2, 3] and [1, 1, 1] may easily be misinterpreted as position points in an ordinary Cartesian coordinate system.
The video actually interprets them as coefficients in front of each sine "axis", corresponding to a composite waveform or function, rather than geometric positions on a plane.
The voiceover first states that points in space are uniformly distributed, then says that when computing area, we can view it as points swept by a moving line segment.
The intuitive assumption of 'points uniformly distributed' supports the method 'area = number of points swept'.
The narrator first explains area via line segment translation, then circle area via line segment rotation about an endpoint.
The derivation of circle area directly applies and extends the method 'line segment sweeps points to yield area'.
The narrator says, 'Okay, let us go one step deeper into three-dimensional space,' then uses square translation sweeping points to define volume.
The three-dimensional volume method is a dimensional generalization of the two-dimensional area method.
The narrator first explains rectangular prism via square translation, then cylinder via square rotation about one side, saying 'according to our earlier logic'.
The derivation of cylinder volume directly applies and extends the method 'square sweeps points to yield volume'.
Both the two-dimensional and three-dimensional cases write 'midpoint trajectory = '.
In the three-dimensional part, the narrator explicitly states 'according to our earlier logic'.
Both use the same intuitive strategy of 'midpoint trajectory as average displacement', one applied to area and the other to volume.
After explaining the cylinder, the narrator immediately asks, 'Then what if we cut this square in half?'
From square rotation to cylinder, then from the resulting right triangle rotation to cone and translation to prism.
The comparison between cones and prisms is a follow-up question raised after halving the square cross-section, continuing the earlier idea of rotating or translating sweeping points.
Narrator compares Euler's formula and matrices in high-dimensional spaces
The video demonstrates the natural generalization of rotation matrices from two to three and then to four dimensions, emphasizing the universality and power of matrices in handling transformations in high-dimensional spaces.
The narrator says, "For example, Minkowski's spacetime light cone can be represented using such a matrix."
Matrices, as tools for spatial transformations, are applied to describe the light cone structure of Minkowski spacetime.
The system of equations is progressively rewritten as a matrix equation.
Systems of multivariable linear equations are a direct application of the concept of coefficient matrices.
The inner product of a single row extends to matrix multiplication with multiple rows.
Each row of the matrix-vector product is the inner product of a coefficient vector with the unknown vector; thus, the inner product is a prerequisite concept for understanding the coefficient-matrix form.
The narrator concludes from the matrix equation that matrices transform vectors into vectors, and names this transformation a linear transformation.
The video generalizes concrete coefficient-matrix multiplication to the broader role of matrices as mappings from vectors to vectors.
The grid deformation animation follows immediately after the definition of matrix transformation.
The grid animation is a geometric application of the concept of linear transformations, used to illustrate how a matrix acts on space.
The narrator says, "In general, we regard the space we inhabit as a flat space."
The narrator says that when computing area, we may assume we are computing the set of points swept by a line segment moving.
The screen displays "midpoint trajectory = " and deduces "circle area = πr²".
The narrator says volume may be assumed to be the set of points swept by a square moving.
The screen displays "midpoint trajectory = " and "cylinder volume = rectangular prism volume = πr³".
The screen finally shows "cone volume = prism volume ???".
The narrator twice uses 'taking the midpoint as the average' to handle rotation-sweeping point problems.
The narrator repeatedly asks why the volumes differ.
The narrator discusses the difficulty of expressing Euler's formula in three-dimensional space.
Screen displays rotation matrix 4x4
Screen displays a matrix containing Cosh and Sinh
The narrator explains that matrices were originally used to solve systems of multiple equations.
Covered · Discuss uniform distribution of points in flat space.
Covered · Two-dimensional line segment translation sweeping points to compute rectangle area.
Covered · Two-dimensional line segment rotation sweeping points to compute circle area, and using the midpoint trajectory to convert it into rectangle area.
Covered · Brief transition: the narrator summarizes the logic seamlessly and prepares to enter three dimensions.
Covered · Three-dimensional square translation sweeping points to compute rectangular prism volume.
Covered · Three-dimensional square rotation sweeping points to compute cylinder volume, and using the midpoint trajectory to convert it into rectangular prism volume.
Covered · Cut the square into right triangles, compare cone and prism volumes, and end with a question mark.
Covered · Propose a logical conjecture that the volumes of a cone and a prism are equal, demonstrate the contradiction arising from their differing volume formulas, and assign it as homework.
Covered · Introduce the concept of matrices, illustrating through animated grid transformations that matrices are a powerful tool for studying spatial transformations.
Covered · Display the two-dimensional rotation matrix and Euler's formula, and point out the limitations of Euler's formula when handling rotations in three or higher dimensions.
Covered · Show how a three-dimensional rotation matrix generates a cone, then extend this idea to four-dimensional rotation matrices, emphasizing the universality of matrices.
Covered · Represent the Minkowski spacetime light cone using a matrix involving hyperbolic functions, and elevate the discussion philosophically by quoting dialogue from 'The Three-Body Problem'.
Covered · The scene shifts to a digital rain background, raising suspense about the true nature of matrices and why they are more powerful than Euler's formula—serving as a lead-in to the next segment.
Covered · Digital rain introduces the origin of matrices via an matrix.
Covered · A system of three linear equations is given, and each individual equation is interpreted as the dot product of two vectors.
Covered · The three dot product results are arranged into a vector, and the coefficients are arranged into a matrix.
Covered · The matrix equation is labeled as consisting of the coefficient matrix, the input vector, and the output vector.
Covered · An analogy is drawn to the movie 'The Matrix', while reminding viewers that the mathematical concept is not as magical.
Covered · A two-dimensional grid deforms under the action of multiple matrices, emphasizing uniform scaling of unit distances.
Covered · The idea of using and as coordinate axes is proposed, and this segment ends here.
Covered · The screen displays "horizontal axis: " and "vertical axis: ", and bends the grid into a sine coordinate system.
Covered · The example point [2; 3] is read as , accompanied by a "combined waveform".
Covered · The voiceover suggests adding as the z-axis, and the screen shows a more complex three-dimensional sine coordinate grid.
Covered · The example point [1; 1; 1] is read as .
Covered · The screen places infinite-dimensional vectors alongside , and the voiceover calls it the sine form of the Fourier series.
Covered · The narration infers that "almost all continuous functions" can undergo similar analysis, reasoning that Fourier series can synthesize other functions.
Covered · The video abandons the bent-coordinate-axis visualization, instead stating that continuous functions can be viewed directly as vectors, and introduces the term "Hilbert space".
Covered · The screen shifts to introductions of Hilbert and his students Hermann Weyl and von Neumann, with no new mathematical formulas.
Covered · Display portraits of Hilbert and his students to illustrate the influence of his school.
Covered · Introduce Hilbert's 23 problems and their significance, listing some specific problems.
Covered · Summarize this episode's content (matrices and Hilbert spaces) and preview the next episode.
Covered · Assign homework and conclude the video.
Reviewed subject paths
Continue with related videos in the same subject.
Bilibili9:8.606999999999971
Bilibili1:15.185000000000002
Bilibili2:55.28800000000001